CN1723739B - Audio signal processing method and processing device - Google Patents
Audio signal processing method and processing device Download PDFInfo
- Publication number
- CN1723739B CN1723739B CN200380105395.1A CN200380105395A CN1723739B CN 1723739 B CN1723739 B CN 1723739B CN 200380105395 A CN200380105395 A CN 200380105395A CN 1723739 B CN1723739 B CN 1723739B
- Authority
- CN
- China
- Prior art keywords
- audio signal
- delay
- digital
- filter
- sampling period
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R2203/00—Details of circuits for transducers, loudspeakers or microphones covered by H04R3/00 but not provided for in any of its subgroups
- H04R2203/12—Beamforming aspects for stereophonic sound reproduction with loudspeaker arrays
Abstract
An audio signal processing method and device using a plurality of digital filters (DF0 to DFn) each supplied with an audio signal and a loudspeaker array (10). Outputs of the digital filters (DF0 to DFn) are respectively supplied to loudspeakers (SP0 to SPn) of the loudspeaker array to form sound fields. A predetermined delay time is set for each of the digital filters (DF0 to DFn) so that a point having a greater sound pressure than the surrounding and a point having a smaller sound pressure than the surrounding are formed. Low-pass filter characteristic is given to the frequency response of the digital filters DF0 to DFn. Moreover, the delay time setting resolution is increased by using a pseudo pulse string.
Description
Technical field
The present invention relates to a kind of acoustic signal processing method and equipment that is fit to be applied to home theater etc.
The application requires the priority of Japanese patent application JP2002-332565 that submitted on November 15th, 2002 and the JP 2002-333313 that submitted on November 18th, 2002, and it is incorporated herein by reference in full.
Background technology
As the speaker system that is fit to be applied to home theater, AV (audiovisual) system etc., 233591 and 1993 years of Japanese Patent Application Publication document 1997 30381 in loudspeaker array is disclosed.As typical example, Fig. 1 illustrates one of conventional loudspeaker array.The loudspeaker array of totally representing with reference number 10 is included in a plurality of loud speakers (loudspeaker unit) SP0 to SPn that is provided with in the array.In this loudspeaker array, n=255, each loud speaker have for example some centimetres diameter.Thereby loud speaker SP0 to SPn in fact bidimensional ground arranges planar.Yet, in the following description,, suppose that loud speaker SP0 to SPn is arranged on the horizontal linear for simplicity of illustration and explanation.
Provide audio signal from source of sound SC to delay circuit DL0 to DLn, they will be postponed preset time τ 0 to τ n respectively at this, and the audio signal after the delay offers loud speaker SP0 to SPn by power amplifier PA0 to PAn respectively.Should be understood that and to be described in detail in delay time T 0 to the τ n that offers audio signal among the delay circuit DL0 to DLn subsequently.
Thereby the sound wave that sends from loud speaker SP0 to SPn will be combined providing acoustic pressure to the listener, locate he or she self position regardless of him or she with respect to loud speaker.Therefore, in the sound field that loud speaker SP0 to SPn shown in Figure 1 forms, will be scheduled to acoustic pressure increases some Ptg and predetermined acoustic pressure and reduces a some Pnc and be defined as follows:
Ptg: the listener should be provided sound as much as possible or acoustic pressure should be than increasing more point in the surrounding environment;
Pnc: the listener should be provided the least possible sound or acoustic pressure should be than reducing more point in the surrounding environment.
Usually, in Fig. 2 or system shown in Figure 3, can take the arbitrfary point to increase a some Ptg as acoustic pressure.
More particularly, suppose in system shown in Figure 2, the distance that increases some Ptg from loud speaker SP0 to SPn to acoustic pressure is respectively L0 to Ln, and the velocity of sound is s, and delay time T 0 to the τ n that provides to sound wave in delay circuit DL0 to DLn is provided with as follows in system shown in Figure 2:
τ0＝(Ln-L0)/s
τ1＝(Ln-L1)/s
τ2＝(Ln-L2)/s
…………
τn＝(Ln-Ln)/s＝0
Thereby, will convert sound wave to by loud speaker SP0 to SPn from the audio signal of source of sound SC, these sound waves will send with delay time T 0 to τ n respectively from each loud speaker SP0 to SPn.Therefore, all sound waves increase some Ptg with arriving at acoustic pressure simultaneously, and the acoustic pressure on acoustic pressure increase point Ptg is with projecting environment.
More particularly, in system shown in Figure 2, the distance that increases some Ptg from loud speaker SP0 to SPn to acoustic pressure is different mutually, and this will cause the time delay of a sound wave to another sound wave.Increase some Ptg by corresponding this time delay of compensation among the delay circuit DL0 to DLn sound is focused at acoustic pressure.Should be understood that hereinafter such system to be called " convergence type system ", as long as suitable, also acoustic pressure increased some Ptg and be called " focus " hereinafter.
In system shown in Figure 3, be arranged on delay time T 0 to the τ n that offers sound wave in the delay circuit DL0 to DLn, it will be identical making the phase front from the capable ripple (sound wave) of loud speaker SP0 to SPn, thereby make the sound wave orientation, and get the increase of sensing acoustic pressure and put the direction of Ptg as anticipated orientation.This system also is regarded as a kind of form of distance L 0 to Ln infinitely-great convergence type.Should be understood that hereinafter such system to be called " orthotype system " that the direction with the phase front unanimity of sound wave is called " anticipated orientation " hereinafter.
In loudspeaker array 10, suitably being provided with of delay time T 0 to τ n forms focus Ptg on the arbitrfary point of permission in sound field, and at same each sound wave of direction interior orientation.And in above-mentioned convergence and orthotype system, because on other any position except a Ptg, the output of loud speaker SP0 to SPn is not in phase made up, they are the most at last by on average, and acoustic pressure will reduce.In addition, in these systems, the voice output of loudspeaker array 10, in case by the metope reflection, may be focused on the Ptg, and point to some Ptg.
Yet above-mentioned loudspeaker array 10 is mainly by using delay time T 0 to τ n convergence or directed sound wave to realize an acoustic pressure increase point Ptg.The amplitude that offers the audio signal of loud speaker SP0 to SPn will only change acoustic pressure.
Therefore, can utilize the directivity of loudspeaker array to be reduced in acoustic pressure and increase acoustic pressure on the some Ptg.For this reason, for example,, perhaps, can rearrange loudspeaker array 10 in order in the direction that reduces some Pnc towards acoustic pressure, to detect less than sound in order when reducing secondary lobe, on the direction of acoustic pressure increase point Ptg, to form main lobe.
For this reason, essential by increasing the quantity n of loud speaker SP0 to SPn, make the size of whole loudspeaker array compare enough big with wave length of sound.Yet in fact this be very difficult to realize.Otherwise the change of acoustic pressure will influence the acoustic pressure that sound wave is assembled and is directed to increases some Ptg.
And, for home theater and AV system etc., essential consideration multichannel sterego.That is, because DVD player is more and more universal, the multichannel sterego source of sound is constantly increasing.Thereby the user should provide the loud speaker with the sound channel as much.Yet so many loud speaker is installed will need sizable space.
And, under the situation that does not reduce tonequality, postponing the audio signal that provides from source of sound SC in order to make delay circuit DL0 to DLn, each delay circuit DL0 to DLn must be made of digital circuit.Particularly, delay circuit can be made of digital filter.In fact, in many AV system, be digital signal because source of sound SC is digital device and an audio signal such as DVD player, under many situations, each delay circuit DL0 to DLn will be made of digital circuit.
Yet, if each delay circuit DL0 to DLn is made of digital circuit, the temporal resolution that then offers the audio signal of loud speaker SP0 to SPn will be subjected to the restriction of the sampling period in digital audio and video signals and the delay circuit DL0 to DLn, therefore can not be less than sampling period.Should be understood that when sampling frequency is 48kHz will approximately be 20.8 μ s sampling period, sound wave will be propagated about 7 millimeters in a sampling period.And the audio signal of 10kHz will be delayed a sampling period that is equivalent to 70 degree phase delays.
Therefore, can not be focused at fully on the Ptg from the phase place of the sound wave of each loud speaker SP0 to SPn, the size that causes focus Ptg is that the observed acoustic image of listener will be bigger, perhaps becomes uncertain as the case may be.
And the sound wave phase place will be more even in the optional position except focus Ptg, thereby, the not abundant reduction of expectability acoustic pressure on other position except a Ptg.Therefore, it is big and uncertain that acoustic image will become, and effect will be than normal conditions difference.
Summary of the invention
Therefore, the objective of the invention is to improve and the acoustic signal processing method of novelty and the above-mentioned shortcoming that equipment overcomes prior art by providing a kind of.
Above-mentioned purpose can realize that according to the present invention, the method comprising the steps of by a kind of acoustic signal processing method is provided: provide audio signal in a plurality of digital filters each; The output of a plurality of digital filters offered each loud speaker in a plurality of loud speakers that constitute loudspeaker array to form sound field; Be arranged on the scheduled delay that will provide in each digital filter in a plurality of digital filters, thereby first harmony that forms the projecting environment of acoustic pressure in sound field forces down second point in surrounding environment; With the amplitude characteristic of adjusting a plurality of digital filters low-pass filter characteristic is offered the frequency response of the audio signal on second o'clock.
In above-mentioned acoustic signal processing method according to the present invention, by point that the projecting environment of acoustic pressure is provided the time of delay that will provide in each digital filter being set and the point that acoustic pressure is lower than surrounding environment being set by the amplitude characteristic of adjusting digital filter.
And, by being provided, a kind of acoustic signal processing method realizes above-mentioned purpose, for example a kind of signal processing method with the delayed digital signal scheduled time, according to the present invention, the method comprising the steps of: is that unit is divided into integer part and fractional part with predetermined time of delay with the sampling period of digital signal; Oversampling comprises the impulse response of the time of delay of being represented by the fractional part at least of scheduled delay so that sampled sequence to be provided, and this sampled sequence of owing to sample is to provide the impulse waveform data of sampling period; With the filter coefficient of these impulse waveform data as digital filter is set, and digital signal is offered the digital filter of operating on sampling period.
Above-mentioned acoustic signal processing method realizes that the fractional part of the time of delay that digital filter needs is with the time of delay that delayed digital signal is suitable.
In conjunction with the accompanying drawings to being used to realize the detailed description of optimal mode of the present invention, these purposes of the present invention and other purpose, feature and advantage will become more apparent according to hereinafter.
Description of drawings
Fig. 1 is the schematic block diagram of the loudspeaker array that comprises in the speaker system of using in home theater and AV system etc.
Fig. 2 is illustrated in the schematic block diagram how loud speaker that comprises in the loudspeaker array forms sound field.
Fig. 3 is expression is formed another example of sound field by the loud speaker that comprises in the loudspeaker array a schematic block diagram.
Fig. 4 explains that locational acoustic pressure suitable in sound field increases some Ptg harmony pressure drop low spot Pnc.
Fig. 5 is the plane graph of the reflection of the sound that sends of the loudspeaker array that is provided with in the room that is shown in as the acoustics enclosure space.
Fig. 6 also is that diagram is because the plane graph of the listener's that the sound reflection in the acoustics enclosure space forms virtual image position.
Fig. 7 A to 7C diagram is because the change of the frequency response that the variation of pulse amplitude causes in the digital filter.
Fig. 8 explains amplitude A 0 to An identification and the backwards calculation by " having influenced the coefficient of the sampling in the CN width " of predesignating space composite pulse response Inc.
Fig. 9 explains that a plurality of somes Pnc1 to Pncm are set to acoustic pressure reduces the amplitude A 0 of some Pnc and definite corresponding points Pnc1 to Pncm to An.
Figure 10 is the schematic block diagram according to first embodiment of audio signal processing of the present invention.
Figure 11 is shown in the flow process of the operation of the Audio Signal Processing in the audio signal processing.
Figure 12 is the schematic block diagram according to second embodiment of audio signal processing of the present invention.
Figure 13 is the schematic block diagram according to the 3rd embodiment of audio signal processing of the present invention.
Figure 14 is the schematic block diagram according to the 4th embodiment of audio signal processing of the present invention.
Figure 15 is the plane graph of the 4 sound channel surround sound fields that formed by a loudspeaker array.
Figure 16 is the schematic block diagram of audio signal processing, wherein forms 4 sound channel surround sound fields by a loudspeaker array.
Figure 17 A to 17D explains the false pulse sequence that forms in the preliminary treatment that is used for by loudspeaker array regeneration.
Waveform, gain characteristic and the phase characteristic of the false pulse sequence that Figure 18 A and 18B diagram are used in the present invention.
Waveform, gain characteristic and the phase characteristic of the false pulse sequence that Figure 19 A and 19B diagram are used in the present invention.
Waveform, gain characteristic and the phase characteristic of the false pulse sequence that Figure 20 A and 20B diagram are used in the present invention.
Waveform, gain characteristic and the phase characteristic of the false pulse sequence that Figure 21 A and 21B diagram are used in the present invention.
Figure 22 is the schematic block diagram according to the 6th embodiment of audio signal processing of the present invention.
Figure 23 is the schematic block diagram according to the 7th embodiment of audio signal processing of the present invention.
Figure 24 is the schematic block diagram according to the 8th embodiment of audio signal processing of the present invention.
Embodiment
At first, will summarize the present invention.In the present invention, because the voice output of the loud speaker that comprises in the next comfortable loudspeaker array of combination in the space is interpreted as pseudo-digital filter to provide response signal on a plurality of points with these points.By predicting that the response signal from " point of alap acoustic pressure should be provided to the listener " does not change the delay that provides to each loud speaker with the amplitude that changes sound, to form the mode control frequency characteristic of digital filter.
By the control frequency characteristic, reduction should provide acoustic pressure on the position Pnc of alap acoustic pressure to the listener, and increases the frequency band that can reduce acoustic pressure.And, reduce acoustic pressure as far as possible naturally.
In addition, according to the present invention, use the frequency oversampling higher to represent the impulse response that postpones, and use the resolution higher to represent this impulse response than the sampling period of this system than the sampling frequency of this audio signal processing.Use the sampling frequency of this system to owe data in the sampling pulse so that the sequence that comprises a plurality of pulses to be provided, and store up this pulse train in databases.When digital audio and video signals being postponed τ 0 to τ n, will be provided for digital filter in the data of databases storage.Because this processing can be provided with the time of delay with temporal resolution higher than the unit delay time precision of the sampling frequency definition of passing through system, can be controlled at the response on the acoustic pressure increase point Ptg harmony pressure drop low spot Pnc more accurately.
Then, will analyze loudspeaker array 10.
For the simplification that illustrates and explain, constitute by n horizontal on straight line loud speaker SP0 to SPn at this hypothesis loudspeaker array 10, this loudspeaker array 10 is configured to as shown in Figure 2 convergence type system.
At this, suppose that each delay circuit DL0 to DLn of convergence type system is made of FIR (finite impulse response (FIR)) digital filter.And, suppose as shown in Figure 4, represent the filter coefficient of Finite Impulse Response filter DL0 to DLn respectively with CF0 to CFn.
And, suppose pulse is offered each Finite Impulse Response filter DL0 to DLn, and on a Ptg and Pnc, measure the output sound of loudspeaker array 10.Should be understood that this measurement is to use the sampling frequency of the regenerative system that comprises digital filter DL0 to DLn or uses the frequency higher than this systematic sampling frequency to carry out.
Subsequently, each response signal of measuring on a Ptg and Pnc will be to send the resulting and signal of acoustics addition with the sound of spatial transmission from all loud speaker SP0 to SPn.At this, in order to understand following explanation better, the output signal of supposing loud speaker SP0 to SPn is the pulse signal that is postponed by digital filter DL0 to DLn respectively.Should be pointed out that and to be called the response signal that after spatial transmission, is superimposed " response of space composite pulse " hereinafter.
Because the delay component of each digital filter DL0 to DLn is set so that voice output is focused on the Ptg, the space composite pulse response Itg that measures on a Ptg will be big pulse shown in Figure 4.And as shown in Figure 4, frequency response (amplitude part) Ftg of space composite pulse response Itg will be smooth on whole frequency, because time waveform adopts the form of pulse.Therefore, acoustic pressure will increase on a Ptg.
Should be understood that, although because the frequency characteristic of each loud speaker SP0 to SPn, the reflection characteristic of the change of spatial transmission process medium frequency characteristic, the wall that in the sound transmission path, exists, by the sampling frequency definition the time base skew or the like, in fact space composite pulse response Itg will not be any pulse accurately, but in order to simplify explanation, will be with desirable model representation at this.To describe in detail hereinafter by sampling frequency definition the time base skew.
The combination of the pulse of base information when the space composite pulse response Inc that will measure on a Pnc on the other hand, is considered as carrying respectively.As can be seen from Figure 4, space composite pulse response Inc is the signal with the pulse that disperses in certain scope.Although should be pointed out that as shown in Figure 4, the impulse response Inc on a Pnc is equally spaced pulse train, and the interval between the pulse train normally at random.Because the relevant information of the position of some Pnc and be not included in each filter coefficient CF0 to CFn and also all original filter coefficient CF0 to CFn all based on direct impulse, so the frequency response Fnc of space composite pulse response Inc also is that all are the combination of the pulse of direct impulse.
Therefore, conspicuous as the design principle according to Finite Impulse Response filter, frequency response Fnc will be smooth in the low frequency frequency range, and the high more then decay of frequency is big more, and as shown in Figure 4, promptly it will have the characteristic that is similar to low pass filter.At this, because the space composite pulse response Itg that increases on the some Ptg in acoustic pressure is big pulse, and the space composite pulse impulse response Inc on a Pnc has the signal that disperses pulse, will be at the frequency response Ftg that is lower than on the level on a Ptg at the frequency response Fnc on the Pnc.Therefore, acoustic pressure will reduce on a Pnc.Hypothesis space composite pulse response Inc is the space Finite Impulse Response filter, Finite Impulse Response filter Inc is made up of the pulse amplitude values sum of the time coefficient that comprises filter coefficient CF0 to CFn at first, by changing the content (amplitude, phase place etc.) of filter coefficient CF0 to CFn, can change frequency response Fnc.That is,, can change the frequency response Fnc that reduces the acoustic pressure on the some Pnc in acoustic pressure by changing filter coefficient CF0 to CFn.
As mentioned above, by form the filter coefficient CF0 to CFn that each delay circuit DL0 to DLn and selection are respectively applied for each digital filter by Finite Impulse Response filter, can acoustic pressure increase point be set and reduce some Ptg and Pnc on the appropriate location in sound field.
Then, will be explained in the interior loudspeaker array of enclosure space.
To the situation of loudspeaker array shown in Figure 3, sound field is an open space at Fig. 1.Yet sound field is as shown in Figure 5 the space RM by wall WL acoustics sealing normally.In this room RM, by selecting the focus Ptg or the anticipated orientation of loudspeaker array 10, the sound A tg that sends from loudspeaker array 10 can be focused on the listener LSNR after by the wall WL reflection around listener LSNR.
In this case, although loudspeaker array 10 is positioned at the place ahead of listener LSNR, will hear sound from behind.Yet, in this case, must be provided with from the sound A tg at rear so that this sound is heard on the highland as far as possible, because it is the sound of expection, sound A nc must be set so that this sound is heard in the lowland as far as possible, because it is unexpected " leakage sound ".
Therefore, as shown in Figure 6,, consider the virtual image in whole room in conjunction with the order of reflection of sound A tg.Because the virtual image can be considered as be equal to Fig. 2 or open space shown in Figure 3, to increase on the position of the virtual image that some Ptg corresponding virtual position Ptg ' are arranged on listener LSNR with acoustic pressure, and the focus or the anticipated orientation of loudspeaker array 10 will be arranged on the position of Ptg ' point.And, acoustic pressure is reduced some Pnc be arranged on the position of actual listener LSNR.
Use the said structure of audio signal processing, virtual speaker can be arranged on the back and the side of multichannel sterego system, thereby realize surround sound regeneration, and needn't loud speaker be set in back and the side of listener LSNR.
Should be pointed out that in order to realize a kind of like this convergence type virtual speaker system according to purpose, application, source of sound content or the like, focus Ptg can be arranged on wall WL and go up or be arranged on any other place, rather than on the position of listener LSNR.And, can not be separately according to the acoustic pressure difference assess technically localization of sound promptly hear sound from direction, will be very important but in this system, increase acoustic pressure.
Then, the acoustic pressure that how to be reduced on the Pnc will be explained.
When as shown in Figure 5 and Figure 6, when listener LSNR is positioned at room RM (enclosure space), also will locate acoustic pressure increase point Ptg so that will determine to depend on the time of delay of filter coefficient CF0 to CFn.When the listener LSNR of location, also locate acoustic pressure and reduce some Pnc, also will determine to reduce the position (the space composite pulse response among Fig. 7 A is identical with space composite pulse response Inc shown in Figure 4) of the pulse of the space composite pulse response Inc that occurs on the some Pnc shown in Fig. 7 A in acoustic pressure.And when the amplitude A 0 to An that changes from the pulse of digital filter DL0 to DLn, controllable sampling width (pulse number) will be the sampling width C N shown in Fig. 7 A.
Therefore, by changing amplitude A 0 to An, the pulse of (in the sampling width C N) shown in Fig. 7 A can change over for example pulse shown in Fig. 7 B of level distribution (response of space composite pulse) Inc ', and frequency response can change over frequency response Fnc ' from frequency response Fnc, shown in Fig. 7 C.
That is to say, reduce the acoustic pressure general who puts on the Pnc in acoustic pressure and only be lowered for the dashed area of the frequency range shown in Fig. 7 C.Therefore, in example shown in Figure 5, will be from the leakage sound Anc of front less than expection sound A tg from the back, thereby will hear sound better from the back.
Even when pulse being changed over space composite pulse response Inc ' by change amplitude A 0 to An, increasing space composite pulse response Itg on the some Ptg and frequency response Ftg in acoustic pressure also will be only be changed for the amplitude of change like this, and can keep the uniform frequency characteristic, this is extremely important.Therefore, according to the present invention, changing amplitude A 0 to An provides frequency response Fnc ' to reduce in acoustic pressure on the some Pnc.
Then, will explain how to determine space composite pulse response Inc '.
The method of determining essential space composite pulse response Inc ' according to space composite pulse response Inc will be explained.
Usually, in order to constitute low pass filter, some methods for designing of window functions such as, Kai Ze peaceful and Blacknam have been recommended to use such as Hamming, the Chinese by Finite Impulse Response filter.The characteristics of the frequency response of the filter by the design of any method in these methods are sharper cut-off characteristics, and this is well-known.In this case, because only the CN sampling can have the pulse duration of using amplitude A 0 to An control, will use window function design low pass filter at this.When the shape of determining window function and sampling counting CN, also will determine the cut-off frequency of frequency response Fnc '.
Count the concrete numerical value that CN determines amplitude A 0 to An according to window function and sampling.For example, by predesignating space composite pulse response Inc " having influenced the coefficient of the sampling in the CN width " as shown in Figure 8, can discern with backwards calculation amplitude A 0 to An.In this case, because a plurality of coefficients will according to circumstances influence a pulse in the space composite pulse response Inc, if the number of corresponding coefficient (number of=loud speaker SP0 to SPn) is less, the situation that does not have coefficient correlation shown in for example as Fig. 8 will be had.
The window width that should be pointed out that window function should preferably be counted the distribution window of CN near equaling to sample.And if a plurality of coefficient has any influence to a pulse in the space composite pulse response Inc, then it is enough to distribute a plurality of coefficients.In the method that this coefficient distributes, preferably should preferentially adjust space composite pulse response Itg is produced less influence and space composite pulse response Inc ' is produced any one amplitude of very big response, yet not stipulate at this.
In addition, the acoustic pressure that a plurality of somes Pnc1 to Pncm can be set to as shown in Figure 9 reduces some Pnc, and uses simultaneous equations to determine that the amplitude A 0 of corresponding points Pnc1 to Pncm is to An.And if if do not meet simultaneous equations or influence amplitude A 0 to the An corresponding points Pcn1 to Pncm not of the concrete pulse of space composite pulse response Inc, as shown in Figure 8, then can use methods such as least square method to determine amplitude A 0 to An, so that they are with the curve of render target window function.
And, can make filter coefficient CF0 to CF2 corresponding to a Pnc1, make filter coefficient CF3 to CF5, make filter coefficient CF6 to CF8 corresponding to a Pnc3 corresponding to a Pnc2, ..., filter coefficient CF0 to CFn perhaps can be set with mutually nested relation and put Pnc1 to Pncm.
In addition, by considering sampling frequency, loudspeaker unit quantity and spatial configuration, can design audio signal processing, the coefficient that wherein influences each pulse of space composite pulse response Inc exists as far as possible at random.And, come implementation space composite pulse response Inc because constitute the space of continuous sequence by the sound that wherein sends together from loud speaker SP0 to SPn, so with the same in the discretization in measuring process, any specific coefficient will can not influence each pulse technically.Yet, for the ease of calculating, explain that at this this system will influence each pulse as coefficient only, verified by experiment as the present inventor, this will can not produce any practical problem.
Then, will be described in detail with reference to the attached drawings preferred embodiments more of the present invention.
First embodiment is the application of the present invention on audio signal processing.Figure 10 illustrates the example of audio signal processing.In Figure 10, diagram is used for the audio signal circuit of a sound channel.That is, (HPF) 11 offers Finite Impulse Response filter DF0 to DFn with digital audio and video signals from source of sound SC by variable high-pass filter, and by power amplifier PA0 to PAn the output of Finite Impulse Response filter DF0 to DFn offered loud speaker SP0 to SPn respectively.
In this case, because can respond the cut-off frequency of the sampling width C N estimated frequency response Fnc ' of Inc, so control the cut-off frequency of variable high-pass filter 11 in conjunction with the cut-off frequency of frequency response Fnc ' according to the inspectable space composite pulse.Under this control, only the audio signal that is in the frequency range that frequency response Ftg is dominant to frequency response Fnc ' of tolerance frequency is passed through.Under situation as shown in figure 11, for example, when the low frequency part of frequency response Fnc ' has the identical level of low frequency part with frequency response Ftg, the effective band of control source of sound, and do not use this low pass part, thus, can only export effective frequency band when hearing sound from behind.
And digital filter DF0 to DFn is included in respectively in the above-mentioned delay circuit DL0 to DLn.In addition, in power amplifier PA0 to PAn, the digital audio and video signals that is provided amplifies its power after through D/A (digital-to-analogue) conversion or the amplification of D class, offer loud speaker SP0 to SPn subsequently.
In this case, in control circuit 12, for example carry out routine 100 as shown in figure 11, and the characteristic of high pass filter 11 and digital filter DF0 to DFn is set as mentioned above.Just, when being provided a Ptg and Pnc, control circuit 12 starts its routine 100 in step 101.Subsequently, in step 102, delay time T 0 to the τ n that will provide in digital filter DF0 to DFn is provided control circuit 12.Then, in step 103, control circuit 12 simulations are counted CN at the space composite pulse response Inc that acoustic pressure reduces on the some Pnc to predict controllable sampling.
Subsequently, in step 104, control circuit 12 calculates can be according to the cut-off frequency of the low pass filter of window function preparation.In step 105, control circuit 12 is listed each effective breadth in the amplitude A 0 to An corresponding with the sampling difference in the pulse train of space composite pulse response Inc, and definite amplitude A 0 is to An.Subsequently, in step 106, according to the result of aforesaid operations, delay time T 0 to the τ n that control circuit 12 is provided with the cut-off frequency of variable high-pass filter 11 and will provides in digital filter DF0 to DFn withdraws from this routine 100 in step 107 subsequently.
By aforesaid operations, control circuit 12 can determine that acoustic pressure increases and reduce some Ptg and Pnc.
Then, will describe the second embodiment of the present invention in detail.
In system shown in Figure 12, for data about the cut-off frequency of variable high-pass filter 11 and delay time T 0 to the τ n that will provide in digital filter DF0 to DFn are provided for a plurality of somes Ptg and Pnc, and with this storage database in the memory cell 13 of control circuit 12.When the data that will be used for a Ptg and Pnc in the operating process at regenerative system offer memory cell 12, extract corresponding data from memory cell 13, and delay time T 0 to the τ n that the cut-off frequency of variable high-pass filter 11 is set and will in digital filter DF0 to DFn, provides.
Then, will describe the third embodiment of the present invention in detail.
In system shown in Figure 13, identical with above-mentioned first embodiment, for example, handle the digital audio and video signals that provides from source of sound SC by variable high-pass filter 11 and digital filter DF0 to DFn.Signal after so handling is offered loud speaker SP0 to SPn through digital addition circuit 14 and power amplifier PA0 to PAn.
In addition, the digital audio and video signals that will provide from source of sound SC and the output of variable high-pass filter 11 offer digital subtraction circuit 15, and it will provide the digital audio and video signals component (illustrated flat among Fig. 7 C) of intermediate frequency and low frequency subsequently.These intermediate frequencies and low frequency digital audio signal are offered digital addition circuit 14 by treatment circuit 16.
Therefore, can correspondingly be controlled at acoustic pressure with the processing of in treatment circuit 16, carrying out and reduce the leakage sound of putting on the Pnc.
Then, will describe the fourth embodiment of the present invention in detail.
Figure 14 schematically illustrates the equivalent electric circuit of the operation that is used for FIR (finite impulse response (FIR)) digital filter DF0 to DFn.As shown in the figure, source of sound SC offers original Finite Impulse Response filter DF0 to DFn with digital audio and video signals by stationary digital high pass filter 17, and the output of digital filter DF0 to DFn is offered digital addition circuit 14.In addition, will offer treatment circuit 16 through wave digital lowpass filter (LPF) 18 from the digital audio and video signals of source of sound SC.
Therefore, under the situation that treatment circuit 16 can be made of digital filter, can carry out its operation by digital filter DF0 to DFn.
Then, will describe the fifth embodiment of the present invention in detail.
Figure 15 and Figure 16 illustrate left front, right front, the left back and right back virtual speaker SP how a loudspeaker array 10 is implemented in listener LSNR
LF, SP
RF, SP
LBAnd SP
RBTo constitute quadraphony surround sound field.
As shown in figure 15, loudspeaker array 10 is arranged on the place ahead of listener NSNR in the room RM.And, as shown in figure 16, so dispose left front sound channel, so that left front digital audio and video signals D
LFTo draw from source of sound SC, and by variable high-pass filter 12
LFOffer Finite Impulse Response filter DF
LF0To DF
LFnThe output of Finite Impulse Response filter is offered loud speaker SP0 to SPn by digital addition circuit AD0 to ADn and power amplifier PA0 to PAn.
And, so dispose right front channels, so that right front digital audio and video signals D
RFTo draw from source of sound SC, and by variable high-pass filter 12
RFOffer Finite Impulse Response filter DF
RF0To DF
RFnThe output of Finite Impulse Response filter is offered loud speaker SP0 to SPn by digital addition circuit AD0 to ADn and power amplifier PA0 to PAn.
In addition, also dispose left back and right back sound channel similarly with left front and right front channels.In Figure 16, use LF that only will be used for left front and right front channels and the reference symbol that RF is replaced by LB and RB to represent these sound channels, therefore will no longer be described at this.
As describing, the numerical value of each sound channel is set with reference to Figure 10 and Figure 14.For left front and right front channels, realize virtual speaker SP by the system of for example describing with reference to figure 1
LFAnd SP
RFFor left back and right back sound channel, realize virtual speaker SP by the system of for example having described with reference to figure 5
LBAnd SP
RBTherefore, these virtual speakers SP
LFTo SP
RBConstitute quadraphony surround sound field.
Because each said system can realize not needing the broad space of the essential usually numerous loud speakers of installation around the multichannel sterego system by a loudspeaker array 10.And, because, therefore do not need the loud speaker that adds only by using the additional character filter just can increase the quantity of sound channel.
In the above embodiment of the present invention, use window function to respond the design principle of Inc ' so that sharper low-pass filter characteristic to be provided as the space composite pulse.Yet, adjust the filter coefficient amplitude by using any other function except window function, also can obtain desired low-pass filter characteristic.
And in the above-described embodiments, filter coefficient is set to all have the pulse train of forward amplitude, so that all space composite pulse responses all are the pulse train with forward amplitude.Yet acoustic pressure reduces some Pnc can be had by the pulse amplitude in each filter coefficient and be set to the characteristic that forward or negative sense define, and keeping lag characteristic simultaneously increases on the some Ptg sound is focused at acoustic pressure.
In addition, in the above-described embodiments, use pulse basically, yet this is the simplification in order to explain as delay cell.Have the basic delay cell of tap conduct of a plurality of sampling of certain frequency response by employing, can guarantee identical effect.For example, delay cell can be the false pulse sequence that guarantees pseudo-oversampling effect basically.In this case, the negative component of amplitude direction is also included within the coefficient, but we can say that so negative unit is similar to pulse on effect.Should be understood that hereinafter and will describe the false pulse sequence in detail.
And, in the above-described embodiments, represent the delay that provides to digital audio and video signals with filter coefficient.Yet this expression also can be applied in the system that comprises delay cell and digital filter.In addition, can increase point and reduce some Ptg and Pnc is provided with one of at least combination or a plurality of combination of amplitude A 0 to An for acoustic pressure.And, for with for example in the realization of virtual rear loud speaker shown in Figure 6 the same fixation application so dispose loudspeaker array 10 so that can expect under the situation of common pip and listening point etc., filter coefficient can be and the acoustic pressure that can be contemplated to increases point and reduces some Ptg and the CF0 to CFn that fixes that Pnc is corresponding.
In addition, in the above-described embodiments, by simulation such as the influence of the SATT that causes in communication process hollow conductance, because the parameters such as phase change that the reflection of reverberation causes, the amplitude A 0 that can determine the filter coefficient corresponding with space composite pulse response Inc ' is to An.And, can measure each parameter in these parameters by suitable measurement mechanism, thereby determine that more suitably amplitude A 0 is to An for simulation more accurately.
And in the above-described embodiments, loudspeaker array 10 comprises the loud speaker SP0 to SPn that arranges along horizontal linear.Yet loud speaker SP0 to SPn also can arrange in the plane or on depth direction.And loud speaker SP0 to SPn also can not necessarily arrange in order.And each the foregoing description all is the convergence type system.Yet the orthotype system can similarly handle.
Then, the delay operation of using false pulse will be explained.
In the above embodiment of the present invention,, is set the time of delay of the unit delay time that defines based on the using system sampling frequency for each digital filter in order to simplify explanation.Yet, more preferably, should be with higher precision setting time of delay.
The pulse train (impulse response) that realization has the time of delay of the temporal resolution more much higher than the unit delay time of using system sampling frequency definition will be called " false pulse sequence " hereinafter.
At first, will explain how to prepare database.
In following explanation, with the symbol that uses as give a definition:
Fs systematic sampling frequency
Nov is the numerical value that temporal resolution is divided sampling frequency 1/Fs.That is, the oversampling frequency is with respect to the multiple of sampling frequency Fs.
Nps by sampling frequency be Fs a plurality of pulses the oversampling cycle 1/ (Fs * Nov) the time base on the number of pulses of approximate representation pulse shape.That is, the quantity of pulse in the false pulse sequence also is the number of times of realizing the digital filter of desired delay.
Example:
Fs＝48kHz，Nov＝8，Nps＝16
At first, the preliminary treatment of the sound reproduction of carrying out for loudspeaker array 10 prepares the false pulse sequence as described above, and is registered in the database.
Just, the preparation database, as describing hereinafter:
(1) according to essential temporal resolution hypothesis oversampling multiple Nov and the pulse number Nps in the false pulse sequence.At this, will explain the Nov increase doubly of temporal resolution, shown in Figure 17 A and Figure 17 B from M pulse to next (M+1) individual pulse.And, sampling period 1/Fs the time duration of Nps pulse is set on the base.
(2) because the oversampling multiple is Nov, Nov oversampling pulse will be included in the cycle of the individual pulse of a M pulse to the (M+1), shown in Figure 17 B.By following formula is set:
m＝0，1，2，......，Nov-1
The oversampling pulse will be taken at time the position (M+m/Nov) on the base of 1/Fs sampling period.Otherwise the oversampling pulse will be taken at oversampling cycle 1/ (time the position (M+Nov * m) on the base of Fs * Nov).
(3) the oversampling pulse in (2) is owed to sample sampling frequency Fs to determine the false pulse sequence, shown in Figure 17 C from sampling frequency Fs * Nov.
In this case, for example, each sequence in (2) can be transformed on the frequency axis by FFT, will be except the frequency inverted of the effective value of the sampling frequency Fs that only owes to sample base then by contrary FFT.And, owe sampling because can be accomplished in several ways, comprise the design of frequency overlapped-resistable filter, will not describe at this and owe sampling techniques.
(4) after this, the false pulse sequence (sequence of Nps pulse) that will determine in above-mentioned (3) is treated to time the pulse on the time location (M+m/Nov) on the base at 1/Fs sampling period virtually.In this case, sampling period 1/Fs the time base on, numerical value M is an integer, numerical value m/Nov is a decimal.
(5) numerical value M is considered as offset information, m/Nov is considered as index information with numerical value, and the corresponding form of related data of the false pulse sequence waveform of determining with these information with in (4) registers in the database 20, shown in Figure 17 D.
Waveform, gain characteristic and the phase characteristic of the middle false pulse sequence that forms in above-mentioned (1) to (4) that Figure 18 to 21 is shown in.Should be understood that Figure 18 to 21 diagram works as Nov=8, these waveforms, gain characteristic and phase characteristic when Nps=16 and m=0 to 7.
For example shown in Figure 18 A under the situation of m=0, the numerical value of time-base waveform is 1.0 in the 8th sampling, is 0.0 in other sampling.So Figure 18 A diagram causes the transmission characteristic of the delay of 8 sampling periods (8/Fs) simply.Along with numerical value m increases, the peak of time-base waveform moves on to the 9th sampling gradually, and this will find out from Figure 18 to Figure 21.At this moment, although frequency gaining characteristic almost is smooth, as finding out from Figure 18 to Figure 21, along with numerical value m increases, the frequency plot characteristic provides bigger phase delay.Just, (the delay of the temporal resolution of Fs * Nov) that by using sampling frequency Fs filtering, realizes having 1/.
The essential preliminary treatment of sound reproduction has been described hereinbefore.The sound reproduction that uses the information in the database 20 will be described hereinafter.
Be used to the sound reproduction of loudspeaker array 10 as the database 20 that in the preparation of above-mentioned database is handled, prepares, as describing hereinafter:
Just, by loudspeaker array 10 regeneration sound, as describing hereinafter:
(11) in series provide digital filter with delay circuit DL0 to DLn.Use digital filter that is provided time of delay, and as their coefficient will be set subsequently with describing.
(12) at first, determine delay time T 0 to the τ n corresponding, and multiply by sampling frequency Fs delay time T 0 to τ n is converted to " having postponed the sampling counting " on the frequency axis of sampling frequency Fs with the position (or anticipated orientation) of focus Ptg.Each delay time T 0 to τ n can be such numerical value, and it has the decimal that can not be represented by the resolution of delay circuit DL0 to DLn.Just, delay time T 0 to the τ n and the counting that postponed to sample are not any integral multiples of the resolution of delay circuit DL0 to DLn.
(13) then, the sampling of delay that will determine in above-mentioned (12) is counted and is divided into integer part and fractional part (fractional part), and time of delay of providing in each delay circuit DL0 to DLn is provided integer part.
(14) subsequently, judge the fractional part of determining of counting that postpones to sample is similar to which the index information m/Nov that accumulates in database 20 in above-mentioned (12).Just, judge this fractional part be similar to 0/Nov, 1/Nov, 2/Nov ..., among (Nov-1)/Nov which.Should be pointed out that if determine that fractional part is similar to Nov/Nov=1.0, then integer part is added 1, and fractional part is defined as being similar to 0/Nov.
(15) from database 20, take out the waveform correlation data of corresponding false pulse sequence according to judged result in above-mentioned (14), and be set to be used for the filter coefficient of the Finite Impulse Response filter of above-mentioned (11).
Use aforesaid operations, the total delay time that offers audio signal by delay circuit DL0 to DLn and digital filter will comprise delay time T 0 to the τ n that determines as in above-mentioned (12).Therefore, in the convergence type system, will assemble on the Ptg of focal position, and locate acoustic image clearly from the sound that loud speaker SP0 to SPn sends.And in the orthotype system, anticipated orientation will be by position Ptg, thereby, also will locate acoustic image clearly.
And because from the sound of loud speaker SP0 to SPn homophase more accurately on focus Ptg, and phase place will alter a great deal on the position except focus Ptg, thereby, on the position except focus Ptg, can reduce acoustic pressure more.Thereby, can more clearly locate acoustic image.
Strictly speaking, owe the sampling techniques except using certain, temporal resolution does not increase in all frequency bands, obtains any very high temporal resolution in the high frequency band with being difficult in.Yet, consider acoustic pressure and the difference between the acoustic pressure on the position except focus Ptg (or unexpected direction) at focus Ptg (or anticipated orientation), in fact can be directed more effectively in nearly all frequency band with apparent sound.
Then, will describe the sixth embodiment of the present invention in detail.
Figure 22 diagram is according to the example of sound reproduction equipment of the present invention.As shown in the figure, sequentially digital audio and video signals is offered digital delaying circuit DL0 to DLn and Finite Impulse Response filter DF0 to DFn, and the output of filter is offered power amplifier PA0 to PAn respectively from source of sound SC.
In this embodiment, be as the integer part in above-mentioned (13) time of delay that provides in each delay circuit DL0 to DLn.And, by as the coefficient of Finite Impulse Response filter DF0 to DFn is set in above-mentioned (15), filter is provided with fractional part time corresponding as described in above-mentioned (13) postpone.In addition, in each power amplifier PA0 to PAn,, make the digital audio and video signals that is provided, offer the corresponding loud speaker among the loud speaker SP0 to SPn subsequently through digital-to-analogue conversion and power amplification or the amplification of D class with such order.
And, preparation database 20.As in the above-mentioned steps that is used for preparing database (1) to (5), preparation database 20, it comprise offset information M and index information m/Nov with as in above-mentioned (4) form of corresponding relation between the waveform correlation data of definite false pulse sequence.According to as in the fractional part search database 20 described in above-mentioned (13), and Search Results is provided for Finite Impulse Response filter DF0 to DFn.And, as the time of delay that is set in the integer part described in (13) in delay circuit DL0 to DLn, provide.
Use is according to the said structure of sound reproduction of the present invention system, promptly be used in delay time T 0 to the τ n that on a Ptg, assembles sound (perhaps being used to make anticipated orientation pass through a Ptg) needs and surpass the resolution of delay circuit DL0 to DLn, also can realize surpassing the fractional part of resolution the time of delay that in each Finite Impulse Response filter DF0 to DFn, provides.
Therefore, under the situation of convergence type system, the sound that sends from loud speaker SP0 to SPn is focused on the focus Ptg, and locatees acoustic image clearly.And under the situation of orthotype system, anticipated orientation also will be located acoustic image clearly by the position of a Ptg.
Then, will describe the seventh embodiment of the present invention in detail.
Figure 23 diagram is according to sound reproduction equipment of the present invention.As shown in the figure, Finite Impulse Response filter DF0 to DFn is also as delay circuit DL0 to DLn.In this embodiment, according to index information m/Nov search database 20.For each Finite Impulse Response filter DF0 to DFn offset information M is set according to Search Results, and, the waveform correlation data of index information m/Nov is set therefore for each filter is provided the time of delay that provides in each delay circuit DL0 to DLn.
Therefore, equally in this sound reproduction equipment, because focus Ptg or anticipated orientation correctly are set, so can locate acoustic image clearly.
Then, will describe the eighth embodiment of the present invention in detail.
Figure 24 diagram is according to sound reproduction equipment of the present invention.This is a kind of form of sound reproduction equipment shown in Figure 23, and wherein digital filter DF0 to DFn will realize such as acoustics such as balanced, amplitude (volume) and reverberation.Therefore, external data in convolution circuit CV0 to CVn with the data convolution of extracting from database 20, and the output of convolution circuit CV0 to CVn is provided for Finite Impulse Response filter DF0 to DFn respectively.
Certainly, also not only be applied to loudspeaker array 10 according to delay of the present invention.For example, this delay being applied to the sound channel used in multidirectional speaker system divides device and allows for the position that woofer and tweeter are adjusted virtual source of sound subtly.Just, can realize so-called time alignment.And, delay according to the present invention can be used in the configuration of in the high definition audio reclaim equiment that uses SACD, DVD-audio frequency etc. hyperfrequency loud speaker depth direction with the desired adjustment of millimeter as unit.
And, in this embodiment, can in memory, calculate in advance and register such as ROM etc., perhaps can be when needed data in the calculated data storehouse 20 in real time.
And, for the speed that is reduced in calculated datas in the database 20, calculate the resource that needs or the data volume in memory, this sound reproduction equipment can so be provided with focus and the anticipated orientation that is not used in other so that will be used for some focus Ptg and anticipated orientation in the data in the database 20.For example, can under the situation that does not have any problem, focus Ptg be positioned at the side of listener LSNR, even setting accuracy is lower than the accuracy that focus Ptg is positioned at listener LSNR the place ahead.Therefore, a kind of so automatic control of not using the data in the database 20 or reducing the number of pulses Nps in the false pulse sequence will allow to limit total data volume and complexity of calculation.
In addition, can automatically change numerical value of N ov and number of pulses Nps according to position and anticipated orientation or complexity of calculation and the hardware capabilities of focus Ptg in each case.And, for example, can strengthen continuously the position of focus Ptg and anticipated orientation etc. dynamically, the effect of real time altering.And, in this case, can dynamically change numerical value of N ov and Nps.
Hereinbefore, described some preferred embodiment of the present invention with reference to the accompanying drawings as an example in detail.Yet, those of ordinary skill in the art is to be understood that the present invention is not restricted to these embodiment, under the situation that does not break away from the scope and spirit of in claims, setting forth and defining, can make amendment in every way, alternatively construct or implement with various other forms.
Industrial applicability
As describing hereinbefore, for the sound of regenerating by loudspeaker array, according to the acoustic pressure of audio signal processing raising of the present invention on desired location, be reduced in the acoustic pressure on the assigned position, and should reduce the position of acoustic pressure and the impulse response of direction be multiply by the spatial window function with synthetic video. Therefore, can reduce wherein can feel easily sound wave from the intermediate frequency of direction (location) and the response in the high-frequency range. At this moment, unnecessary extensive increase loudspeaker array this means that system according to the present invention is very practical.
And, in order to set up the multichannel sterego field, can use single loudspeaker array to realize that around the multichannel sterego field this is exclusively used in narrow space loudspeaker is installed.
In addition, by adopting the false pulse sequence is set each time delay, can be set the time delay that resolution ratio is lower than unit delay time. Thereby focus and anticipated orientation are very clear and definite, thereby will locate clearly acoustic image. And, what being lower than focus and anticipated orientation on its point because acoustic pressure is in office, this also will be conducive to the expliciting the position of acoustic image.
Claims (20)
1. acoustic signal processing method may further comprise the steps:
Provide audio signal in a plurality of digital filters each;
The output of a plurality of digital filters offered each loud speaker in a plurality of loud speakers that constitute loudspeaker array to form sound field;
The scheduled delay that will provide in each digital filter in a plurality of digital filters is provided, makes audio signal arrive at propagation delay time of first in the sound field with consistent with each other by each digital filter and each loud speaker; With
Adjust the amplitude characteristic of a plurality of digital filters, low-pass filter characteristic is offered the synthetic response of the audio signal on second in sound field.
2. according to the acoustic signal processing method of claim 1, wherein make sound wave after by the metope reflection, arrive in first and second at least one from loudspeaker array.
3. according to the acoustic signal processing method of claim 1, wherein, when in sound field, forming and at first at second, by calculating the filter coefficient of each digital filter in definite a plurality of digital filters, and this filter coefficient is set for each digital filter.
4. according to the acoustic signal processing method of claim 1, wherein, when in sound field, forming and at first at second, read the filter coefficient of each digital filter a plurality of digital filters from database, and be that each digital filter in a plurality of digital filters is provided with this filter coefficient.
5. according to the acoustic signal processing method of claim 1, wherein:
Be unit the sampling period with audio signal, will be divided into integer part and fractional part for the scheduled delay of at least one setting in described a plurality of digital filters;
Oversampling comprises the impulse response of the time of delay of representing with the fractional part at least of scheduled delay on the cycle shorter than sampling period, and so that sampled sequence to be provided, and this sampled sequence of owing to sample is to provide the impulse waveform data of sampling period; With
Described impulse waveform data are set to the filter coefficient of digital filter, so that described digital filter provides the time of delay corresponding with described fractional part.
6. according to the acoustic signal processing method of claim 5, wherein, with delayed audio signal part integral multiple sampling period, scheduled delay, and its delay is comprised remainder described fractional part, scheduled delay by the digital delaying circuit of on sampling period, working by digital filter.
7. according to the acoustic signal processing method of claim 5, wherein:
The oversampling cycle of described oversampling operation is the 1/N of the sampling period of described audio signal, and N is the integer more than or equal to 2; With
When time of delay of representing by described fractional part during, adopt m/N as described fractional part, wherein m=1 to N-1 near integer m times of described oversampling cycle.
8. according to the acoustic signal processing method of claim 7, wherein:
Postpone to be stored in advance in the database as the impulse waveform data of time of delay of the m/N of sampling period; With
From the impulse waveform data of being stored, take out the impulse waveform data that are similar to described fractional part, and it is set to the filter coefficient of digital filter.
9. according to the acoustic signal processing method of claim 1, wherein:
Be unit the sampling period with audio signal, will be divided into integer part and fractional part for the scheduled delay of at least one setting in described a plurality of digital filters;
Oversampling comprises the impulse response of the time of delay of representing with the fractional part at least of scheduled delay on the cycle shorter than sampling period, and so that sampled sequence to be provided, and this sampled sequence of owing to sample is to provide the impulse waveform data of sampling period; With
Transmission characteristic convolution in the impulse waveform data of predetermined acoustical effect is provided, and is set to the filter coefficient of digital filter as the synthetic waveform data of convolution results.
10. an audio signal processor comprises a plurality of digital filters, and each digital filter is provided with audio signal, wherein:
Each digital filter in a plurality of digital filters each loud speaker in a plurality of loud speakers that constitute loudspeaker array provides output to constitute sound field;
Each digital filter in a plurality of digital filters has scheduled delay, makes audio signal arrive at propagation delay time of first in the sound field with consistent with each other by each digital filter and each loud speaker; With
Each digital filter in a plurality of digital filters has amplitude characteristic, low-pass filter characteristic is offered the synthetic response of the audio signal on second in sound field.
11., wherein make sound wave after by the metope reflection, arrive in first and second at least one from loudspeaker array according to the audio signal processor of claim 10.
12. audio signal processor according to claim 10, wherein, when in sound field, forming and at first at second, by calculating the filter coefficient of each digital filter in definite a plurality of digital filters, and this filter coefficient is set for each digital filter.
13. audio signal processor according to claim 10, wherein, when in sound field, forming and at first at second, read the filter coefficient of each digital filter a plurality of digital filters from database, and be that each digital filter in a plurality of digital filters is provided with this filter coefficient.
14. according to the audio signal processor of claim 10, wherein:
Be unit the sampling period with audio signal, will be divided into integer part and fractional part for the scheduled delay of at least one setting in a plurality of digital filters;
A counting circuit also is provided, be used for comprising that by oversampling on the cycle shorter the impulse response of the time of delay of representing with the fractional part at least of scheduled delay is to provide sampled sequence than sampling period, and this sampled sequence of owing to sample, calculate the impulse waveform data of sampling period; With
The impulse waveform data that provided by counting circuit are set to the filter coefficient of digital filter.
15. according to the audio signal processor of claim 14, wherein:
The oversampling cycle of described oversampling operation is the 1/N of the sampling period of described audio signal, and N is the integer more than or equal to 2; With
When time of delay of representing by described fractional part during, adopt m/N as described fractional part, wherein m=1 to N-1 near integer m times of described oversampling cycle.
16. according to the audio signal processor of claim 10, wherein:
Be unit the sampling period with audio signal, will be divided into integer part and fractional part for the scheduled delay of at least one setting in described a plurality of digital filters;
A counting circuit also is provided, be used for comprising that by oversampling on the cycle shorter the impulse response of the time of delay of representing with the fractional part at least of scheduled delay is to provide sampled sequence than sampling period, and this sampled sequence of owing to sample, calculate the impulse waveform data of sampling period; With
The transmission characteristic that the predetermined acoustical effect is provided in the impulse waveform data convolution with the synthetic waveform data that are provided as convolution results filter coefficient as digital filter.
17. according to the audio signal processor of claim 10, wherein:
Be unit the sampling period with audio signal, will be divided into integer part and fractional part for the scheduled delay of at least one setting in a plurality of digital filters;
Storage device also is provided, is used to store by oversampling on the cycle shorter and comprises that the impulse response of the time of delay of representing with the fractional part at least of scheduled delay is to provide sampled sequence and the impulse waveform data of the sampling period that this sampled sequence of owing to sample provides than sampling period; And
Taking-up is in the impulse waveform data of described storage device stored, and it is set to the filter coefficient of digital filter.
18. according to the audio signal processor of claim 17, wherein:
The oversampling cycle of described oversampling operation is the 1/N of the sampling period of described audio signal, and N is the integer more than or equal to 2; With
When time of delay of representing by described fractional part during, adopt m/N as described fractional part, wherein m=1 to N-1 near integer m times of described oversampling cycle.
19. according to the audio signal processor of claim 18, wherein:
Postpone to be stored in advance in the described storage device as the impulse waveform data of time of delay of the m/N of sampling period; With
From the impulse waveform data of in described storage device, storing in advance, take out the impulse waveform data that are similar to described fractional part, and it is set to the filter coefficient of digital filter.
20. according to the audio signal processor of claim 10, wherein:
Be unit the sampling period with audio signal, will be divided into integer part and fractional part for the scheduled delay of at least one setting in described a plurality of digital filters;
Storage device also is provided, is used to store by oversampling on the cycle shorter and comprises that the impulse response of the time of delay of representing with the fractional part at least of scheduled delay is to provide sampled sequence and the impulse waveform data of the sampling period that this sampled sequence of owing to sample provides than sampling period; And
The transmission characteristic that the predetermined acoustical effect is provided in the impulse waveform data convolution with the synthetic waveform data that are provided as convolution results filter coefficient as digital filter.
Applications Claiming Priority (5)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP332565/2002 | 2002-11-15 | ||
| JP2002332565A JP3821228B2 (en) | 2002-11-15 | 2002-11-15 | Audio signal processing method and processing apparatus |
| JP333313/2002 | 2002-11-18 | ||
| JP2002333313A JP3951122B2 (en) | 2002-11-18 | 2002-11-18 | Signal processing method and signal processing apparatus |
| PCT/JP2003/013082 WO2004047490A1 (en) | 2002-11-15 | 2003-10-10 | Audio signal processing method and processing device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1723739A CN1723739A (en) | 2006-01-18 |
| CN1723739B true CN1723739B (en) | 2011-08-03 |
Family
ID=32697556
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200380105395.1A Expired - Fee Related CN1723739B (en) | 2002-11-15 | 2003-10-10 | Audio signal processing method and processing device |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JP3821228B2 (en) |
| CN (1) | CN1723739B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| RU2719325C2 (en) * | 2015-09-08 | 2020-04-17 | Сэн-Гобэн Изовер | Method and system for obtaining at least one acoustic parameter of environment |
Families Citing this family (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4349123B2 (en) * | 2003-12-25 | 2009-10-21 | ヤマハ株式会社 | Audio output device |
| JP2005197896A (en) * | 2004-01-05 | 2005-07-21 | Yamaha Corp | Audio signal supply apparatus for speaker array |
| JP4161906B2 (en) | 2004-01-07 | 2008-10-08 | ヤマハ株式会社 | Speaker device |
| JP4251077B2 (en) * | 2004-01-07 | 2009-04-08 | ヤマハ株式会社 | Speaker device |
| JP3915804B2 (en) | 2004-08-26 | 2007-05-16 | ヤマハ株式会社 | Audio playback device |
| JP2006210986A (en) * | 2005-01-25 | 2006-08-10 | Sony Corp | Sound field design method and sound field composite apparatus |
| JP4779381B2 (en) | 2005-02-25 | 2011-09-28 | ヤマハ株式会社 | Array speaker device |
| JP4949638B2 (en) * | 2005-04-14 | 2012-06-13 | ヤマハ株式会社 | Audio signal supply device |
| JP4189682B2 (en) | 2005-05-09 | 2008-12-03 | ソニー株式会社 | Speaker check device and check method |
| FI122089B (en) * | 2006-03-28 | 2011-08-15 | Genelec Oy | Calibration method and equipment for the audio system |
| KR100717066B1 (en) * | 2006-06-08 | 2007-05-10 | 삼성전자주식회사 | Front surround system and method for reproducing sound using psychoacoustic models |
| KR100765793B1 (en) | 2006-08-11 | 2007-10-12 | 삼성전자주식회사 | Apparatus and method of equalizing room parameter for audio system with acoustic transducer array |
| JP4984825B2 (en) * | 2006-10-27 | 2012-07-25 | ヤマハ株式会社 | Sound generation apparatus and sound generation system |
| KR100788702B1 (en) | 2006-11-01 | 2007-12-26 | 삼성전자주식회사 | Front surround system and method for reproducing sound using beam forming speaker array |
| JP2008252625A (en) * | 2007-03-30 | 2008-10-16 | Advanced Telecommunication Research Institute International | Directional speaker system |
| JP4975112B2 (en) * | 2008-01-31 | 2012-07-11 | 三菱電機株式会社 | Band division time correction signal processing apparatus |
| KR101295849B1 (en) | 2008-12-18 | 2013-08-12 | 삼성전자주식회사 | Apparatus for controlling sound directional radiation pattern and method thereof |
| JP5486568B2 (en) * | 2010-12-21 | 2014-05-07 | 日本電信電話株式会社 | Audio spot reproduction processing method, apparatus, and program |
| JP5486567B2 (en) * | 2010-12-21 | 2014-05-07 | 日本電信電話株式会社 | Narrow-directional sound reproduction processing method, apparatus, and program |
| JP2017163432A (en) * | 2016-03-10 | 2017-09-14 | ソニー株式会社 | Information processor, information processing method and program |
| US10089063B2 (en) * | 2016-08-10 | 2018-10-02 | Qualcomm Incorporated | Multimedia device for processing spatialized audio based on movement |
| US20190387320A1 (en) * | 2016-12-28 | 2019-12-19 | Sony Corporation | Audio signal reproduction apparatus and reproduction method, sound pickup apparatus and sound pickup method, and program |
| CN108449689A (en) * | 2018-05-09 | 2018-08-24 | 广州市迪霸电子音响有限公司 | A kind of speaker for realizing ultrahigh frequency |
| US10623882B1 (en) * | 2019-04-03 | 2020-04-14 | xMEMS Labs, Inc. | Sounding system and sounding method |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4817149A (en) * | 1987-01-22 | 1989-03-28 | American Natural Sound Company | Three-dimensional auditory display apparatus and method utilizing enhanced bionic emulation of human binaural sound localization |
-
2002
- 2002-11-15 JP JP2002332565A patent/JP3821228B2/en not_active Expired - Fee Related
-
2003
- 2003-10-10 CN CN200380105395.1A patent/CN1723739B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4817149A (en) * | 1987-01-22 | 1989-03-28 | American Natural Sound Company | Three-dimensional auditory display apparatus and method utilizing enhanced bionic emulation of human binaural sound localization |
Non-Patent Citations (1)
| Title |
|---|
| JP特开平8-191225A 1996.07.23 |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| RU2719325C2 (en) * | 2015-09-08 | 2020-04-17 | Сэн-Гобэн Изовер | Method and system for obtaining at least one acoustic parameter of environment |
Also Published As
| Publication number | Publication date |
|---|---|
| JP2004172661A (en) | 2004-06-17 |
| JP3821228B2 (en) | 2006-09-13 |
| CN1723739A (en) | 2006-01-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1723739B (en) | Audio signal processing method and processing device | |
| EP1562403B1 (en) | Audio signal processing method and processing device | |
| AU2020203222B2 (en) | Generating binaural audio in response to multi-channel audio using at least one feedback delay network | |
| CN101194536B (en) | Method of and system for determining distances between loudspeakers | |
| KR100608025B1 (en) | Method and apparatus for simulating virtual sound for two-channel headphones | |
| JP3565908B2 (en) | Simulation method and apparatus for three-dimensional effect and / or acoustic characteristic effect | |
| JP4177413B2 (en) | Sound reproduction apparatus and sound reproduction system | |
| RU2747713C2 (en) | Generating a binaural audio signal in response to a multichannel audio signal using at least one feedback delay circuit | |
| EP1341399B1 (en) | Sound field control method and sound field control system | |
| EP1881740B1 (en) | Audio signal processing apparatus, audio signal processing method and program | |
| US20040131338A1 (en) | Method of reproducing audio signal, and reproducing apparatus therefor | |
| TW201727623A (en) | Apparatus and method for sound stage enhancement | |
| JPWO2006004099A1 (en) | Reverberation adjustment device, reverberation correction method, and sound reproduction system | |
| CN102334348B (en) | Converter and method for converting an audio signal | |
| US6738479B1 (en) | Method of audio signal processing for a loudspeaker located close to an ear | |
| EP3090573B1 (en) | Generating binaural audio in response to multi-channel audio using at least one feedback delay network | |
| Wierstorf et al. | Assessing localization accuracy in sound field synthesis | |
| CN112005492B (en) | Method for dynamic sound equalization | |
| CN109076302A (en) | Signal processing apparatus | |
| Stephenson | Assessing the quality of low frequency audio reproduction in critical listening spaces | |
| Townsend | Enhancements to the generalized sidelobe canceller for audio beamforming in an immersive environment | |
| GB2366975A (en) | A method of audio signal processing for a loudspeaker located close to an ear | |
| KR100705930B1 (en) | Apparatus and method for implementing stereophonic | |
| Rosen et al. | Automatic speaker directivity control for soundfield reconstruction | |
| Roper | A room acoustics measurement system using non-invasive microphone arrays |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20110803 Termination date: 20151010 |
|
| EXPY | Termination of patent right or utility model |