TW293229B - Filter network reverberation generator - Google Patents

Abstract

A filter network reverberation generator for processing the audio signals, and which includes: 1. a plurality of wave-ladder filters, each of them simulated the audio signals sent or reflected by wave transmission from a certain direction,input the original audio signals into the wave-ladder filter and created different oscillation in each direction; 2. At least one of the wave-ladder filter will contain one or several secondary space distribution matrix to simulate the middle reflex object; 3. An adder which combined with the output signals of all the wave-ladder filters to create the reverberation signals.

Description

Printed by the Employee Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs ^ 3229 Λ7 Β7 V. Description of the invention (I) Background of the invention This invention relates to a reverb generator of sound signals. Reverberation usually produces a sound with a reflective object The reason for it in the space is that it is caused by the original wave and the multiple reflections combined by the reflected waves from the floor, walls and other objects. Generally speaking, in an acoustic space that sounds pleasant, there are energy samples that decay like an exponential function. Since the human auditory system is more accustomed to being in a reverberant environment, recordings in a reverberant environment usually also sound less natural. Today, many recordings in the studio have added some damping effect (damping effect), so that the reverb effect is almost gone. In order to make the recording result sound like it is produced in a real space (such as a stadium), some post-processing methods are used, and the artificial reverberation is the most common among them. Artificial reverberation (artificial reverberation) devices must produce the original signal plus a natural reverberation in the acoustic environment. A pleasant reverb device must meet at least the following criteria. The sound energy pattern must be attenuated exponentially. The mid-frequency portion of the sound spectrum usually contains longer reverb than the low and high frequencies. The reflection in all directions must be irregular in terms of intensity and length of time. This system must deal with early reflection and long reverberation according to the above-mentioned rules. This paper music scale is applicable to the A4 specification (210 XW7 mm) of the Central Circle Standard (CNS) (Please read the back first Note to fill out this page) •-°

T A7 B7 printed by the Beigong Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economy V. Description of the invention (>) Traditional artificial reverb technology uses mechanical devices or electronic circuits to simulate mechanical devices, but with digital signal processing (Digital Signal Processing ) With the rapid development of technology, modern artificial reverberation devices are mostly digital. Among the many good processing methods, one way to produce a pleasing reverb is to measure the impulse response (irnpulse response) of a good acoustic space and then use this response to produce a finite impulse response (Finite Impulse Response (FIR)) and the original signal to do a round-robin operation ( convolution) produces output. Usually, the number of filters (taps) is as few as a few thousand, and as many as tens of thousands. This kind of calculation is very considerable. However, in most applications, the practice of this method is very impractical, so 'too many digital reverb generators are a combination of two methods. One is early reflection (early reflection) uses a finite impulse response, and Long reverberation is expressed in Infinite Impulse Response (IIR) in order to further reduce the amount of calculation. However, when using an infinite impulse response filter (IIR filter) in the reverberator, some difficulties arise. First, it produces a comb filter effect that is annoying to the ear. Again, near the end of the reverberation, the 'sound tends to become louder, and sometimes harsher. Third, if any filter is located in the feedback loop of an infinite impulse response filter (IIR), the stability issue is an intractable issue. Finally, the characteristics of infinite impulse response filters to simulate the characteristics of sound is another problem. Summary of the Invention One of the reasons why traditional methods are difficult to produce high-quality reverberations' is that the basic models of these technologies do not consider the acoustic environment at all. This paper scale is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) (Please read the precautions on the back before filling out this page), -0 A7 B7 293229 V. Description of invention (Xi) Go in, therefore, the above artificial reverberation (artificial reverberation :) method is stimulated by a solid space to produce Reverb. Therefore, we established a reverb technology based on imaginary acoustic room. Generally speaking, this technique is based on a digital wave-ladder filter to simulate the propagation of one-dimensional sound waves. A virtual space is based on the geometric distribution of this space, which consists of several digital wave ladder filters, and simulates the situation of re-distribution of various vectors at the intersection of acoustic waves at various positions. The digital wave ladder filter and the spatial distribution matrix (Spatial Distribution) ^ 311 ^^ (801 ^)) form an artificial reverb generator, and the stability of this reverb generator can also be easily achieved, as long as these Filters and spatial distribution matrices are lossy from an energy point of view. This system is also very easy to adjust the system parameters without affecting the stability, such as the size of the virtual space and the gain of the reverberation. To produce real early reflections, we used finite impulse response (FIR) filters to simulate the signals before they entered the system filter network. 'Generally speaking, in the concept of this invention, it is just a collection of filter networks. The basic components of the filter network are derived from the one-dimensional waveform propagation model, and the spatial distribution matrix (SDM ) Is used to connect various basic components. The structure of this network is used to simulate the propagation of sound waves in a virtual acoustic space. The sound signal is input into a finite impulse response filter (FIR) to obtain early reflection. As a result, after inputting this filter network to obtain a long reverberation, the effect of reverb is generated, and this filter The stability of the network can also be proved. This paper scale is applicable to the Chinese National Standard (CNS) Α4ϋ ^ (210X297mm) (please read the notes on the back before filling in this page) 2. Description of the invention (屮) Generally speaking, each element of this invention is known to those in the field of digital signal processing and spatial acoustics. The imaginary room, ladder filter and spatial distribution matrices are also known to those who specialize in this field. Therefore, this invention combines these known techniques and techniques to obtain good reverberation (such as the sound of sweetness, stability, elasticity, etc.). Looking at the present invention from another angle, as long as the principle of wave propagation is used, a reverb generator that sounds pleasant can be easily generated. The invention can produce an artificial reverb that sounds more natural and substantial than the traditional way. It can be applied to the processing of sound signals, and can also be applied to the setting of home and commercial environments, such as theaters, concert halls, karaoke, etc. Additional benefits and features will become more apparent due to the more specific description and scope of patent applications proposed below. Brief description of the legend Figure 1 is a block diagram of a discrete-time system that generates a one-dimensional wave propagation. Printed and printed by the Beigong Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs (please read the notes on the back to fill in this page) Figure 2 is a block diagram of a signal reflection and transmission at a junction. Figure 3 is a block diagram of one-dimensional wave conduction with multiple reflection intersections in a discrete-time mode. Figure 4 is a block diagram of a one-dimensional wave conduction practiced in discrete-time mode with multiple reflection junctions, plus the use of an attenuator to simulate a lossy signal path. This paper uses the Chinese National Standards (CNS) Α4 specification (21 Οχ 297 public) A7 B7 _ 5. Description of the invention (5) Figure 5 is a τ-dimensional wave with multiple reflection junctions practiced in discrete time Conduct and use a low-pass filter in the signal path to simulate a block diagram of an attenuated signal path. Figure 6 is a block diagram of a finite impulse response (FIR) filter to simulate 'early reflection. Figure 7 is a block diagram of an all_pass filter practiced with an infinite impulse response (IIR). Figure 8 is a block diagram of a comb filter practiced with an infinite impulse response (IIR). Figure 9 is a block diagram of a filter network reverb generator using a finite impulse response (FIR) filter 'followed by four wave ladder filters ° Figure 10 is a digital filter network reverb generator and it The functions supported by the side, and this is the practice of Figure ll.b. Figure Π.a and Figure 1 l.b are examples of virtual spaces that can be created by the filter network reverb generator. '-------- {install-- (please read the precautions on the back and then fill in this page), vs i; the detailed description of the invention printed by the Consumer Cooperative of the Central Standard Falcon Bureau of the Ministry of Economic Affairs. The embodiment is used as a reverb generator for post-processing of audio recording. The input sound signal is divided into two stages for processing: 'Finite Impulse Response (FIR) filter to simulate early reflection' and then a filter network to simulate long reverberation. The reverb generator is 'that is' based on the simulation of sound wave propagation. It converts continuous time (c ο ntin U 〇us -1 ime) -dimensional wave conduction into discrete time that can be practiced with the digital filter 10 of Fig. 1. model. Simulating wave conduction in digital form is known to those in the field of digital signal processing.-This paper standard is applicable to the Chinese National Standard (CNS) A4 specification (210X297 public jitter A7 B7 29S22d V. Description of the invention (6) And there are many books on this aspect that can be used as a reference 'such as LR Rabiner et al., Digital Processing of Speech Signals, Prentice-Hall, Inc., 1978. Figure 1 shows a simulated one-dimensional wave conduction without loss (Lossless) digital wave ladder filter 'it simulates the direct reflection of an object' like a wall '. The input is the original audio, and the output signal is the sound that returns to the original signal position after being reflected. This simple digital wave ladder filter includes Simulate the sound waves of the upper branch (upper branch) in the outward direction, and the lower branch (lower branch) in the direction of the sound wave reflected from the wall. Each branch is established by several delay modules 16 (delay module) The number of modules in each branch and the amount of delay in each module are determined by the size of the virtual acoustic space we want to simulate. A gain module 18 (ga in module) After connecting from the upper branch 12 to the lower branch M, it is equivalent to the amount of sound reflected back from the wall. Usually the amount of reflection will be less than 1, so the gain of the gain module will also be 1. The gain module The delay in the gain and delay module ^ in the program are programmable, so the user can adjust various parameters to match the characteristics of the sound. Since the sound wave can be moved from the end to the end_fine movement, This situation can be simulated in Figure 2 and Figure 3. Figure 2 shows that the lossless digital wave ladder has one segment-it is modeled by the empty_distribution matrix 24 (s as a model-dimensional wave_shenma Na, although there is no __ in the picture, the far _ wall _ seems to be represented by the upper part _ lower part of the right side of the reading path connected to the 1 benefit module. It is described in the previous figure. In this example, use Four sets of delay modules 20 (1), 2) are used, and the size of this paper is remotely used by China National Standard (CNS & A4 specifications (210X297mm) (please read the precautions on the back before filling out this page). The Central Falcon Bureau Beigong Consumer Cooperative printed the Ministry of Economic Affairs Central Standards Bureau employee consumption Cooperative cooperative printing A7 B7 5. Description of invention (7) 22 (1) and 22 (2). The delay module 20 (1) represents the transmission of sound waves from a pre-determined position to the space distribution matrix (SDM). The time spent between the represented contacts. The delay module 22 (1) represents the time it takes for the reflected sound waves to return from the spatial distribution matrix (SDM) through the same path as described above to the previously determined position. time. Therefore, the delays of modules 20 (1) and 22 (1) are indeed the same. It can be inferred that all delay modules' if they are located on the same acoustic path segment, they should have the same delay. It is used to simulate a sound wave passing through the junction of the same door, and then enter a space of different size, so part of the sound wave passes through the junction and part of the sound wave is reflected back. Similarly, the returned sound wave also undergoes the same process at the junction, that is, part of the sound wave returns to the main space through the junction, and part of the sound wave returns to the adjacent space. If the exact situation in the space can be accurately calculated, then the amount of reflection and transmission at each junction can be accurately determined. However, the actual situation in these spaces is quite difficult to simulate and measure. Therefore, a more practical approach is to generate a virtual acoustic space as described above, and then adjust the control parameters, such as the delay and coefficient of the space allocation matrix (SDM), to create the desired acoustic reverb effect. The Spatial Distribution Matrix (SDM) 24 in Figure 2 is simulated with diagonal coefficients (diag ο na 1 c 〇efficients) through the junction and coefficients (〇ff-diag ο na I c 〇efficients) that are not diagonal The case where the sound wave is reflected back from the junction, in this case it is a second-order (2X2) matrix. Figure 5 shows the case where three wave ladder filters as shown in Figure 4 are connected together. A paper scale applies to China National Standard (CNS) A4 specification (210X297mm) --------- <. Pack — (please read the notes on the back first and fill in this book)-Printed by the Staff Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs Install A7 B7. Fifth, the description of invention (8), this combination can be used to simulate multiple partially reflected objects in the acoustic path between the original contact and the wall, and each intermediate object reflects back part of the sound wave. The amount of delay is determined by the size of the virtual space to be simulated. In theory, the reflection coefficient can also be calculated from the characteristics of the virtual space, such as geometry and air density. However, in practice, it is very elastic, but because of this, how to produce good acoustic results is the main choice. Since acoustic energy usually attenuates gradually as it propagates in space, it is practical to add an attenuation factor to the wave ladder filter, as shown in Figures 4 and 5. Both Figures 4 and 5 show a one-dimensional wave conduction reflection represented by a spatial distribution matrix, and a propagation path between junctions plus an object with lossy (lossy). Except for adding a module for simulating energy attenuation in the acoustic path, Figure 3 is similar to Figures 4 and 5. In the example in Figure 4, the branch of the simulation outgoing sound wave is added with gain modules 30 (1), 30 (2) with gains +1, ^ 2, respectively, and the sound waves reflected back are simulated The branch also added gain modules 32 (1) and 32 (2) with gains §1 £ + 1 and §1 £ + 2, respectively. In the example of Figure 5, the filter modules 34 (1) and 34 (2) added to the branch, and the other modules added to the branch are the filter modules 36 (1) and 36 (2) . As shown in Figure 4, the gain modules at the same part of the acoustic path should have the same gain. Therefore, the same attenuation coefficient should be applied to the same part of the acoustic path, but the direction is opposite to the gain module -9 ~ This paper scale is applicable to the Chinese National Standard (CNS) (2 丨 0X2.97mm) (please Read the precautions on the back first and then fill out this page). Binding. Order ^ 93229 A7 B7 5. Description of the invention (closed)-··. In other words, the gain modules 30 (1) and 32 (1) should have the same value, and the gain modules 30 (2) and 32 (2) should also have the same value. In the lossy digital wave ladder filter shown in Figure 5, the filter can be a finite impulse response (FIR) filter as shown in Figures 7 and 8 or an infinite impulse response (IIR) filter . Both of these can be further elaborated and discussed in the current filter design book. Usually low-pass filter (low-pass filter) represents the absorption in the air. As we all know, high-frequency signals usually decay much faster than low-frequency signals. Of course, when designing a virtual space, users can not add The above parameters. However, analog low-frequency signals decay relatively quickly and it is understandable, as long as they sound good. In any design, the user can use any filter to control the attenuation rate. If the above wave ladder filter is applied to simulate early reflection (early reflection), it usually takes too much memory to build up the calculation amount. In order to reduce the calculation amount and memory, the finite impulse response (FIR) shown in Figure 6 ) The filter 40 can be used to simulate early reflections. The delay module 42 usually does not necessarily have the same delay amount and weight G for each delay, as shown in Figure 6, therefore, the output of the filter is a function of the input signal: output = Gmik m + Gm. iik.m + i + ...... + Gi ik.i + G〇ik The concept of early reflection is discussed in the book of Elements of Computer Mu sic by J. Moor. Usually ' Early reflection refers to those that are closer to the source body from the floor, wall or ____ _-10 -______ This paper scale uses the national standard Lee (CNS) A4 specification (210X297 mm) (please read the notes on the back first (Fill in this page again) Printed with the A7 B7 Employee Consumer Cooperative of the Ministry of Economic Affairs. Printed on the A7 B7 Employee Consumer Cooperative of the Central Sample Agency of the Ministry of Economic Affairs. 5. Description of the invention. () Sound waves reflected by other objects. Usually the delay of early reflection The time is distributed between 0 and 100 milliseconds. The determination of the proportion and delay in the finite impulse response filter (FIR filter) is known to those in the field of signal processing. So the details will not be described in detail However, usually the best way Directly measure the impulse response of a good acoustic space such as Carnegie Hall in New York City within 100 milliseconds, the measured impulse response can be used in this finite impulse response (FIR) filter, but this The method will be time-consuming and laborious. Figure 9 is a block diagram of a reverb built by the above digital wave ladder filter. The digitized input signal is processed by the finite impulse response (FIR) filter 50 to produce early reflections. (Early reflection) effect. This filtered signal is then input into the digital wave ladder filter 52 network to produce a long reverberation. In this embodiment, four digital wave ladder filters 52 are included Coupling via the spatial distribution matrix 54. The digital wave staircase filter is built according to the following principles. The spatial distribution matrix 54 combines the signals fed back from the wave staircase filter (for example, the signal labeled P 丨 j is U ..., 4), regenerate the output signal (for example, the signal labeled P 丨 'j is 1, ..., 4), and then combine it with the signal filtered by the finite impulse response (FIR) to generate separate digital waves The input signal of the step filter. The spatial distribution matrix (SDM) 54 calculates the sound wave pressure of the signal returned by each signal path simulated by the digital wave step filter. Then, the spatial distribution matrix (SDM) 54 calculates the output and then The sound wave pressure (eg P :) that is fed back to the digital wave ladder filter. -11--------- {袭 ------ book (please read the precautions on the back and fill in this page) This paper Wave scale General Chinese National Standard (CNS) A4 wash grid (2 丨 0> < 297mm) 293229 A7 B7 V. Description of invention (U) Space allocation matrix (SDM) 54 with four branches, which can be as follows Column matrix operation practice: ^

Printed by the Central Standards Bureau of the Ministry of Economic Affairs, the Consumer Labor Cooperative, where P. , J = l, 2, 3, 4 'represents the signal propagated by the jth path of the space allocation matrix, P 丨, j = l, 2, 3, 4 represents the input of the jth path of the space allocation matrix signal of. In this example, J is a 4x4 matrix, and the size of the space allocation matrix is of course determined by the number of input and output matrices. The coefficients of the space allocation matrix can be derived from some known acoustic theories, or can be arbitrarily specified. value. However, in order to achieve stability, this method requires the magnitude of the eigenvalues of the space allocation matrix to be 1 as the upper limit. The spatial distribution matrix used in the digital wave ladder filter is to simulate the intersection of reflections, and there is also a general notation as shown above. But a typical example has two sets of output and input paths. Therefore, the matrix J is a 2x2 matrix. The output of the finite impulse response (FIR) filter that has been attenuated by 1 / η (η = 4, η is the number of digital wave ladder filters) is added to the output of each spatial distribution matrix 54 by the adder%, resulting in The input of each digital wave ladder filter. Another adder 58, combined with the output sound pressure of the spatial distribution matrix (SDM) 54, forms a reverb signal, and another attenuator 60, with a programmable gain value G, controls the reverb signal Strength of. -· 12-This paper scale is applicable to China National Standard (CNS) A4 (2 丨 0X297mm) -------- ^ · Installation-- (Please read the precautions on the back before filling this page)

-, 1T A7 A7 Printed by the Employee Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economy V. Description of invention (\ 7-) The design and establishment of the number-k wave staircase filter is not so important. However, they are used to simulate a dummy acoustic Space is the key point, and the design of this special digital wave ladder filter is left to the system user to make a heavy decision. Of course, it is determined by some objective factors, such as the quality or characteristics of the system sound. Normally, the sound quality is required, and the system is of course more complicated. In addition, a more complex system gives the user more room to round the parameters, making the reverb generator produce a more suitable sound. Figure 10 is a specific example of a complete reverb generator. This system is designed to simulate a virtual space with acoustic characteristics as shown in Figure 11.b. The number of Fig. 10 is the number of bit wave step filters equivalent to the number of branches used in Fig. 11.b to simulate acoustic space. In this example, there are 12 branches. The audio signal 62 is first sampled by an analog-to-digital converter 64 and becomes the input signal of a finite impulse response (FIR) filter. The central part of this circuit is the same as shown in Figure 9 except that there are U digital wave ladder filters in this example instead of the four in Figure 9. The matrix J of the practice space allocation matrix (SDM), in this example, is a 1 2 X 1 2 matrix. Finally, after the attenuated signal output from the attenuator 60 is combined with the original signal representing the direct sound signal through an adder 68, it is input to a digital / analog converter 70 to generate the final output analog signal. The parameter control unit 72 can be used: a microprocessor or a completed hardware to practice 'it controls the length of the digital wave ladder filter (that is, the amount of delay), the amount of attenuation of the sound wave (such as gain), space allocation Samples and remnants ___-13-This paper scale is applicable to the Chinese national standard falcon (CNS) A4 cash (210X297mm) {-¾ clothing — (Please read the precautions on the back before filling this page)

, 1T 293229 A7 B7 V. Description of Invention (Θ) Printed by the Employee Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs

The amount of attenuation. Changing these parameters is equivalent to changing the acoustic characteristics of the virtual space, but the basic shape of the virtual space remains unchanged. The design program starts with the establishment of a model of virtual space, as shown in Ha. Or ll.b. The last point in these figures is the space allocation matrix designed based on the above-mentioned principles. Represents the digital 'wave ladder wave filter. The end node of each branch can be used to represent the pure reflection of the original signal in the direction of attenuation after attenuation. The model of the virtual space can be designed according to the principles of spatial acoustics described in some books. As published by John Wiley Co. Fundamentals of Acoustics, Spring Hall Verlag Co., Concert Hall Acoustics and Journal of Acoustics Society of America. Some papers published in journals. The virtual space can be of any shape or size, and its acoustic characteristics, such as dimensions, shape or whether it is a lively or dead room, are also selectable. 'The number of delay elements in each digital wave ladder filter is determined by the size of the virtual space branch. Gain factor (gain factor) is usually said to be less than unity. If the filter is added to the wave ladder filter, the combined energy is gradually attenuated in the real space environment. The fact that the low frequency signal is less attenuated than the high frequency signal, so the low pass (l〇w_pass) or band pass (band-pass ;) The filter is more in line with the needs. Also remember a few additional rules are very helpful for designing the model. For example, "If only a few wave ladder filters are used in the system" G (please read the precautions on the back before filling this page)-install '1Τ

A7

The amount of delay should not be too large. In addition, the total amount of delay in each branch should not be an integer multiple of other branches. If this rule is not followed, certain frequencies with the same delay length will be over-rendered and emphasized, which will easily produce unimagined results. As for the coefficient in the space allocation matrix (SDM), it is better to be less than 1 to ensure the stability of the system. In order to achieve a natural and pleasant sound, it is best to let the coefficient be random. However, the effect of this result is not as great as the effect of the aforementioned branch delay length. A simple three-dimensional acoustic space is calculated from the signals reflected by the four walls, floor and ceiling. If each wall represents a branch, then at least six branches are needed, that is, the space allocation matrix (SDM) ) Is a 6x6 matrix. Du Printed by the Ministry of Economy, Central Bureau of Standards and Staff Consumption Cooperation --------- 1¾—— (please read the precautions on the back before filling out this page) An example of a more complex virtual space model is shown in Figure 11 in. In this example, at the junction of the branches, adjacent branches are connected to each other via a space allocation matrix (SDM). Therefore, the Space Allocation Matrix (SDM) 80 couples the two crossover branches to generate signals that are transmitted and reflected back, so the matrix J in the Space Allocation Matrix (SDM) is a 4x4 matrix. However, the eigenvalue in the matrix must still be less than 1. It must be clarified that the above techniques can be used to simulate the real acoustic space. However, comparing with virtual space, it is found that the simulation of real space is more difficult in the design of digital wave ladder filter or in the selection of variables. Although the foregoing embodiment is a digitized practical method, the circuit of the present invention can also be formed by an analog method or a hybrid method. Of course, if you adopt -15-this paper scale is applicable to the Chinese National Standard (CNS) A4 specification (210X297mm) A7 B7 V. Invention description (β) ~ Shanghai '· The analogy method does not require the analogy / digit of Figure 10. And digital / analog conversion: converter. The remaining specific descriptions are included in the following patent scope. Binding line (please read the precautions on the back before filling in this page) Printed by the Ministry of Economic Affairs, Central Standard Falcon Bureau Employee Consumer Cooperatives This paper size is applicable to the Chinese National Standard (CNS) A4 (210X297mm)

Claims (1)

A8 B8, C8, D8, Beigong Consumer Cooperative, Central Standards Bureau, Ministry of Economic Affairs. A. B8, C8, D8. 1. Processing of audio signals: ^: Reverb generator for wave network, this device includes: a plurality of wave ladder filters (wave -ladder filter), each of which is a sound signal that simulates the relative transmission or reflection of the wave conduction in a specific direction. In each of the above-mentioned multiple wave ladder filters, the original sound signal is input and generated in all directions For different oscillations, at least one of the above wave ladder filters must contain one to several secondary space distribution matrix (SDM) modules to simulate objects that are intermediately reflected, and an adder combines all the above wave ladder filters The output signal of the generator is used to produce a reverb signal. 2. For example, the filter network reverb generator of item 1 of the patent scope further includes a primary space distribution matrix (SDM) module, and this module couples the plural wave ladder filters, In other words, receive the reflected signal of the complex wave ladder filter as the input terminal and then generate a plurality of output signals as the input of the wave ladder filter = 3. The filter network as described in item 1 of the patent application scope Road reverb generator, where more than one wave ladder filter is composed of one or more secondary space distribution moment barriers (SDM) that simulate intermediate reflective objects. 4. The filter network reverb generator as described in item 2 of the patent scope, in which more than one wave ladder filter is composed of one or more secondary space distribution moments that simulate an intermediate reflective object Barrier (SDM). The size of this paper is applicable to the Chinese National Standard (CNS) A4 specification (210 X 297 mm) -------- ^ 1 pack ------ order ----- "Line {Please read the back of the first (Note ** Please fill in this page again.) 293229 as C8 D8, the employee consumer cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs. 6. Patent application scope 5. The filter network reverb generator as described in item 1 of the patent application scope, the At least one wave ladder filter may further add one to several gain factors that simulate the attenuation of the reflected sound signal in the digital wave ladder filter. 6. The filter network reverb generator as described in item 3 of the patent application scope, the at least one wave ladder filter can be further added with one to several analogues of the attenuation of the reflected sound signal in the wave ladder filter Gain factor (gain factor) ° 7. The filter network reverb generator as described in item 4 of the patent scope, the at least one wave ladder filter can be further added with one to several analog wave ladder filters Gain factor of internal reflection sound signal attenuation amount 8. The filter network reverb generator as described in items 1, 5, 6, or 7 of the patent application scope, the at least one wave step filter The filter can be further added with one to several filters that simulate the frequency of the sound reflected in the wave-step filter. 9. The filter network reverb generator as described in item 8 of the scope of patent application, in which the oven can be selected from a group of low-pass (low-pass) and bandpass (bandpass) filters. 10. The filter network reverb generator as described in item 2 of the patent application scope, in which the main space allocation matrix and one to several secondary space allocation moments. This paper standard is applicable to the Chinese National Standard (CNS) A4 specification ( 210 X 297 mm) ^ —Binding (line (please read the notes on the back before filling in this page) A8 B8 C8 D8 V printed by the Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs “Patent Application Array” is composed of a group The characteristic value (eigenvalue) whose boundary value is less than 1. U · The filter network reverb generator as described in item 2 of the patent application scope, further includes a finite impulse response (FIR) filter that simulates early reflection of sound 12. The filter network reverb generator as described in item 2 of the patent scope, in which a plurality of wave ladders and the main spatial distribution matrix (SDM) module further contain variable elements To control the transfer characteristics of the network. 13. The filter network reverb generator as described in item 12 of the patent application scope further includes a parameter control unit (parameter control unit) to control multiple variable elements in complex wave ladder filters and the main spatial distribution matrix: 14. The filter network reverb generator as described in item 13 of the patent scope, Among them, a plurality of variable elements (variable elements) include a delay module with a variable delay amount. 1 5. The filter network reverb generator as described in item 13 of the patent application scope, of which a plurality of variable elements Variable elements include variable gain gain modules. -19-I ^ 1! ^. I Binding I 『Each (please read the precautions on the back and then fill in this K)) The paper scale is applicable to China National Standards (CNS) Α4 said grid (210Χ297mm> ABCD 293229 ~, patent application scope 16. The filter network reverb generator described in item 12 of the patent application scope, the main space distribution matrix module includes A plurality of variable coefficients, and these variable coefficients are controlled by a parameter control unit. 17. The filter network reverb generator as described in item 13 of the patent scope, It includes a finite impulse response (FIR) filter that simulates the early reflection of sound. 18. The filter network reverb generator as described in item 17 of the patent application also includes a gain module, which uses a Controllable gain factor (gain factor) to amplify or attenuate the reverb signal. 19. The filter network reverb generator as described in item 18 of the patent application scope further includes a second adder, which combines the reverb signal adjusted with the gain and the original sound signal to generate the output sound signal . 20. The filter network reverb generator as described in item 1 of the patent scope can be constructed digitally. 2 1. The filter network reverb generator as described in item 1 of the patent application scope can be constructed by analogy. This paper scale is applicable to the Chinese National Standard (CNS) A4 (210X297mm) ^-binding (line (please read the precautions on the back before filling in this page) Printed by the Consumer Labor Cooperative of the Central Standardization Bureau of the Ministry of Economic Affairs