JP2000333299A - Surround acoustic signal generator - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reproduce an acoustic characteristic of a surround sound field space with higher fidelity. SOLUTION: The surround acoustic signal generator has a coefficient ROM 6 that stores a coefficient of a frequency versus level characteristic that is obtained by applying FFT arithmetic operation to a sound impulse response characteristic by direction in a plurality of sound environments in advance and FFT computing elements 2a-2e that apply FFT arithmetic operation to an incoming digital audio signal, multiply the arithmetic result by the coefficient selected from the coefficient ROM 6 and apply inverse FFT arithmetic operation to the multiplied data so as to reproduce a surround acoustic space with higher fidelity.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a surround sound reproducing apparatus, and more particularly to an apparatus for reproducing high quality surround sound.

[0002]

2. Description of the Related Art Conventionally, a configuration as shown in FIG. 6 has been used as an example of an apparatus for obtaining a surround reproduction signal. According to this device, input signals Li (left), Ci (center), Ri corresponding to signal sources arranged in stereo for a listener.
(Right), Sli (rear left), Sri (rear right), LFEi (for heavy bass) are input to the surround playback device, and signals with a surround effect added Lo (left), Co (center), Ro (right) ), Slo
(Rear left), Sro (Rear right), and LFEo (for heavy bass) are schematically configured to be output corresponding to each speaker arrangement in a stereo arrangement as shown in FIG. Further, the configuration will be described in detail while explaining its operation. Speaker input signal Li placed in front of the listener,
The buffers Ci and Ri are buffer-amplified by the corresponding buffer amplifiers b, and a first added synthesized signal obtained by adding the respective signals from the adder / synthesizer 10 is obtained.
To obtain a first initial reflected sound generation signal. The buffer amplifier b is a well-known linear amplifier and is a circuit provided to prevent the deterioration of the inter-channel separation characteristic due to the influence of the adder circuit provided at the subsequent stage.
Further, the rear speaker input signals Sli and Sri are supplied to the second adder 11 via the respective buffer amplifiers b to obtain a synthesized signal, which is input to the second initial reflected sound processing unit 14 and the second initial reflected sound processing unit 14 Obtain a reflected sound generation signal. The second initial reflected sound generation signal and the first initial reflected sound generation signal are amplified by the respective buffer amplifiers b and synthesized by the third addition / synthesis unit 16 to obtain a third addition / synthesis signal. On the other hand, the added signal output from the adder 11 is supplied to the adder 12, added to the added signal output from the adder 10, and the synthesized signal is supplied to the reverberant sound generator 15. Then, the reverberation sound generation unit 15
, And the output signal combined by the adder 16 are combined by an adder 17, and these combined signals are supplied to an adder 18, and the input signals Li,
Signals Lo, Ro, Slo, and Sro to which Ri, Sli, and Sri have been added to add a predetermined reverberant sound are provided, and these signals are input as surround signals together with the input Ci and LFEi amplified by each amplifier. Output to the speaker.

FIG. 7 shows the detailed configuration of the reverberation generator 15, which is known as a device devised by Schrader. Note that FIG. 7 shows that the reverberation sound generator 15 has Lo, Ro, Sl
Since four types of signals for o and Sro can be obtained, a plurality of signals corresponding to those are provided, but one of the signal processing circuits is shown. This circuit 15 is roughly composed of four systems of circuits for generating first to fourth reflected sounds indicated by reference numerals 20 to 23 and a reverberation sound adding circuit for generating reverberant sounds indicated by reference numerals 25 and 26. Be composed. The first initial reflected sound generation circuit has delay circuits 20b and 20g having a delay time of τ1.
An amplifier 20c having a feedback gain of 1 and an addition circuit 20a for adding the input signal and the signal of the feedback amplifier 20c. The second to fourth early reflected sound generators have delay times τ2, τ3, τ4. , And is configured to generate initial reflected sounds with different delay times by the same operation. These output signals are added and synthesized by an adder 24 and supplied to a first reverberation component adding circuit 25 in the next stage.

The first reverberant sound adding circuit 25 and the second reverberant sound adding circuit 26 provided at the subsequent stage have the same configuration, and have feedback gains of delay circuits 25b and g5 having a delay time of τ5. It comprises an amplifier 25c, an adding circuit 25a for adding the input signal and the signal of the feedback amplifier 25c, an inverting amplifier 25d having a gain of g5, and a delay circuit 25b and an adding circuit 25e for adding the signal of the inverting amplifier 25d. The second reverberation sound adding circuit 26 has a delay circuit having a delay time of τ6, and is configured to generate reverberation sounds having different delay times by the same operation.

[0005]

By the way, the generation of the initial reflection sound and the reverberation sound in the above-mentioned prior art apparatus is such that the delay time is simply given to the input signal and the attenuation characteristic of the reverberation sound signal is given. There was only. Further, the reverberation sound generation unit as shown in FIG. 7 uses a method based on a combination of a filter circuit having a comb-shaped characteristic and an all-pass filter that allows all signals to pass, and selects the feedback gain of the comb-shaped filter. Real-world sound field characteristics are simply realized by generating reverberation components, etc. In these methods, signals that simulate the sound field with simple parameters such as early reflections and reverberation components The sound field having complicated transmission characteristics could not be precisely reproduced. In addition, the actually measured sound field data is only used for calculating the amplitude level, delay time, and reverberation time of the reflected sound, and has not been effectively used as data for sound field reproduction. Was.

Accordingly, in the present invention, the transfer characteristics of the sound field to be reproduced are measured as impulse response characteristics for each direction to the listener or calculated by sound field simulation, and the input signals (Li, Ci, Ri) of each channel are calculated. , Sl
i, Sri) and the obtained direction-specific impulse response characteristics are convoluted with arithmetic processing, thereby enabling precise sound field reproduction and providing an inexpensive configuration for reproducing them. Moreover, the calculation speed during reproduction can be increased.

[0007]

The present invention comprises the following means 1) and 2) to solve the above-mentioned problems. That is,

1) A surround sound signal generating apparatus for generating a multi-channel surround sound signal for reproducing a sound signal in a selected sound field space on the assumption that a speaker is arranged around a listener. A coefficient storage means for storing coefficients of frequency versus level characteristics obtained by performing FFT operation on impulse response characteristics for each direction of a sound field in a plurality of acoustic environments, and storing sound signal data of an incoming digital audio signal. An arithmetic processing unit for performing an FFT operation, multiplying the data based on the operation result by the coefficient selected from the coefficient storage unit, and performing an inverse FFT transform on the multiplied data. Surround sound signal generator.

2) In a surround sound signal generating apparatus for generating a multi-channel surround sound signal for reproducing a sound signal in a selected sound field space, assuming a speaker arranged around a listener, A coefficient storing means for dividing a plurality of impulse response characteristics for each direction of a sound field in a plurality of acoustic environments, and storing these divided characteristics as coefficients of frequency versus level characteristics obtained by performing FFT operations, respectively; The acoustic signal data of the incoming digital audio signal is subjected to an FFT operation, and each of the divided coefficients of the impulse response characteristics for each direction selected from the coefficient storage means and the incoming signal corresponding to the time length of these coefficients. Multiplied by the data that has been subjected to the FFT operation based on the data, and the multiplied data is subjected to an inverse FFT transform. A surround sound signal generating apparatus, comprising: at least a calculating means for adding the T signals.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of a surround sound signal generating apparatus of the present invention will be described below. FIG. 1 is a schematic configuration of a surround sound signal generation device according to the embodiment. In the figure, Li, Ci, Ri, Sli, Sri, and LEFi, which are analog audio signals generated corresponding to the assumed speaker arrangement, are supplied from respective input terminals. These analog audio signals are converted into digital signals by AD converters 1a to 1f and supplied to FFT calculators 2a to 2e. The FFT calculators 2a to 2e include an input / output buffer for storing input / output audio data of a well-known configuration (not shown) for performing processing to be described later, a buffer for temporary storage for FFT calculation, an FFT calculation processing unit, , An inverse FFT operation processing unit, an addition unit, and the like. On the other hand, a sound field selection command signal is supplied to the coefficient ROM 6 by operating means (not shown).
In advance, in a plurality of acoustic environments to be realized, impulse response characteristics for each direction of the sound field to be reproduced for each speaker,
Alternatively, the coefficient of the frequency versus level characteristic obtained by performing the FFT operation on the impulse response characteristic for each direction is stored. On the output side, there is provided a buffer 7 for extracting and temporarily storing the coefficient selected by the sound field selection command signal. The FFT calculators 2a to 2e perform a convolution operation based on the above-described input analog audio signals and the coefficients called by the buffer 7 based on predetermined processing steps to be described later to perform a desired surround environment. To obtain a characteristic signal corresponding to. And
The surround signal generated by the FFT calculators 2a to 2e and the LFEi signal delayed by a predetermined time by the delay unit 8 are:
The analog signals are converted into analog signals by the DA converters 3a to 3f, respectively, and amplified as output signals for driving the speakers by the amplifiers 4a to 4f. The speakers L, C, R, Sl, Sr, and LEF
It has a general configuration capable of reproducing a sound with reverberation as if it were heard in a desired music hall, for example. In the system in which the signal LFEi is processed, the signal LFEi is a signal for driving a subwoofer that performs heavy bass reproduction, and includes signals in the mid-range and treble related to the direction perception of the audio signal. Therefore, there is no need to perform convolution processing by FFT using the impulse response for each direction. Here 2a ~ 2e
A delay unit 8 for compensating for the delay time of the circuit block is inserted.

Next, referring to FIG. 2, the procedure for creating the coefficients to be stored in the coefficient ROM 6 and the convolution operation of the FFT operation will be described in further detail. FIG. 5 is a flowchart showing these processing steps. Steps 1 to 4 are procedures for creating coefficients to be stored in the coefficient ROM 6. For example, data measured by a personal computer or the like as impulse response characteristics for each direction with respect to a listener or data calculated by simulation is previously calculated. This is the procedure to prepare. First,
In step 1, transfer characteristics of a sound field to be reproduced are measured as impulse response characteristics for each direction to the listener, or surround effect data calculated by sound field simulation is input in the form of impulse response characteristics. The obtained impulse response characteristics are divided in the time axis direction to obtain first divided impulse response data and second divided impulse response data. In step 3, FFT calculation processing is performed on both the first divided impulse response data and the second divided impulse response data obtained by the division processing. In step 4, the first FFT operation is performed.
And the second divided impulse response data are stored at predetermined positions in a coefficient ROM 6 as a memory circuit. The coefficient ROM 6 stores impulse response data obtained by subjecting a large number of transfer characteristics of the sound field to be reproduced to FFT calculation processing at respective predetermined positions, and selects necessary one of the stored impulse response data from the sound field. The transfer characteristics of the selected sound field are selected according to an instruction and used as data to be convolved with the input digital audio signal. In step 5, digital audio signals are obtained from the AD converters 1a to 1e,
In step 6, the obtained digital audio signal is temporarily stored in the above-mentioned input / output buffer (not shown). In step 7, the digital audio signal data stored in the input / output buffer is read out while being divided into the same length as the divided impulse response data, and is stored in the temporary storage buffer for the FFT operation. Step 8
Then, the data of the digital audio signal stored in the temporary storage buffer is obtained, and the FFT operation is performed in the FFT operation unit by a technique such as a butterfly operation to obtain data of frequency versus level characteristics.

In the multiplication processing in the multiplication unit in step 9, the digital audio data on the frequency axis subjected to the FFT operation and the result of the FFT operation of the impulse response data on the frequency axis obtained in step 4 are read out.
The data of the same frequency is multiplied in parallel at the timing described later to obtain a multiplication result. In step 10, the obtained result of the multiplication processing on the frequency axis is subjected to inverse FFT operation processing in the inverse FFT operation processing unit, and data converted to a digital audio signal on the time axis is obtained. Here, there are two pieces of data stored and accumulated in the respective predetermined positions of the coefficient ROM 6 in step 4, ie, the first divided impulse response data and the second divided impulse response data subjected to the FFT operation processing. Is performed on each of the two divided impulse response data to obtain two sets of multiplication processing results. In step 10, the two sets of multiplication results obtained by the multiplication are calculated by the inverse F
FT operation processing is performed to obtain two sets of inverse FFT output signals. In step 11, the obtained two sets of inverse FFT output signals are subjected to addition processing in an adder to obtain a set of digital audio signals to which a surround effect has been added, and temporarily store them in an input / output buffer circuit. The operation of one of the FFT operators 2a to 2e has been described above, but the operation of the other FFT operators is the same. And
In a next step (not shown), a set of digital audio signals to which the surround effect is temporarily stored in the input / output buffer is output from each FFT arithmetic unit, and the surround effect is provided via each DA converter. Output as analog audio signals.

Next, an FF for performing a convolution operation process
The operation timing of the T calculator will be described with reference to the timing chart of FIG. For example, the selected impulse response data h (n) having a length of N samples is divided into data h0 (n) and data h1 (n) with a length of N / 2 samples and stored in the coefficient ROM 6. These data are temporarily stored in the buffer 7 from the coefficient ROM 6 for calculation, and the digital audio signal that is continuously input is the divided response data h0 (n) and h1 (n). X'0 (n), x'1 in the temporary store buffer divided into the same length
(n), x'2 (n), x'3 (n), x'4 (n), x'5 (n), ...
It is stored as In addition, the FFT calculation processing when these are stored is performed in units of time when the number of samples is N / 2.

As shown in FIG. 1, the multiplication process is performed by dividing the system into two systems.
Input signal x 'separated by half the number of samples
0 (n) is h0 (n) and h1 during the period when the next x'1 (n) is input.
Performs convolution operation with (n), h0 (n) * x'0 (n) and h1 (n) *
x'0 (n) is obtained. H0 (n) * x 'during the period when the next x'2 (n) is input
0 (n) and h1 (n) * x'0 (n) each perform an inverse FFT operation to obtain a signal obtained by adding the obtained two signals. In this way, the input signal is subjected to the convolution operation of h0 (n) and h1 (n), the inverse FFT operation, and the addition processing, and outputs a digital audio signal in which the impulse response signal for each direction is convolved.

For example, the timing chart of FIG.
(n) and h1 (n) are timing charts when not divided, and the hardware configuration according to this example has the same configuration as that of FIG. 1 as described above. Since the optimal impulse response in each direction is convolved by FFT operation for each channel, accurate surround sound field reproduction is possible, but the impulse response characteristic described above is further reduced to N / 2 samples in the first half. The delay time is 2N longer than that in the case of dividing into the latter half N / 2 samples.
It is possible to reduce to N samples.

In the above-described embodiment, the apparatus is configured to divide the impulse response data into two. However, the present invention is not limited to this, and the same effect can be obtained with a configuration in which the impulse response data is divided into three and four. Further, in the above embodiment, the input signal is described as an analog signal. However, when a digital signal is input, it can be directly input to the FFT calculator. Further, the output signal of the digital audio is converted to an analog signal by DA conversion, amplified, and supplied to a speaker. When the output signal is supplied to an audio device capable of inputting a digital signal having these functions, an FFT is used. The same effect can be obtained even if the output of the arithmetic unit is used as it is as the output of the surround sound signal generation device.

[0017]

According to the first aspect of the present invention, the reproduction of the acoustic signal in the sound field space is performed by the convolution operation using the impulse response data for each direction of the sound field to be reproduced, which has been subjected to the FFT processing. Since the signal is generated, there is an effect of outputting a high-accuracy surround sound signal. According to the second aspect of the present invention, in particular, since the FFT operation process is performed by dividing the impulse response characteristics for each direction, in addition to the effect of the first aspect, the generation time of the surround sound signal is shortened. effective.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram of a surround sound signal generation device according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating a flow of a convolution operation process used in the surround sound signal generation device.

FIG. 3 is a diagram showing a relationship between arithmetic processing operation timings in the apparatus according to the embodiment of the present invention.

FIG. 4 is a diagram showing a relationship between arithmetic processing operation timings using undivided impulse response characteristics in the device according to the embodiment of the present invention.

FIG. 5 is a diagram showing a positional relationship of a speaker arrangement with respect to a viewer used in the surround sound reproducing device.

FIG. 6 is a block diagram of a conventional surround signal reproducing device.

FIG. 7 is a block diagram of a reverberation sound generation circuit unit used in a conventional surround signal reproduction device.

[Explanation of symbols]

1a, 1b, 1c, 1d, 1e, 1f AD converter 2a, 2b, 2c, 2d, 2e, 2f FFT operator 3a, 3b, 3c, 3d, 3e, 3f DA converter 4a, 4b, 4c, 4d, 4e, 4f Amplifier 5a Front left speaker 5b Front center speaker 5c Front right speaker 5d Rear left speaker 5e Rear right speaker 5f Subwoofer speaker for bass reproduction 6 Coefficient ROM 7 Buffer 8 Delay unit 10, 11, 12 Addition circuit 13 , 14 Early reflection sound processing unit 15 Reverberation sound generation unit 16, 17, 18 Adder 20, 21, 22, 23 Initial reflection sound generation circuit 20a, 21a, 22a, 23a, 24, 25a, 25
e, 26a, 26e, 28 Addition circuit 20b, 21b, 22b, 23b, 25b, 26b Delay circuit 20c, 21c, 22c, 23c, 25c, 26c Feedback circuit 25d, 26d All-pass filter 25, 26 Reverberation sound generation circuit 27 Buffer amplifier b buffer amplifier

Claims (2)

[Claims] 1. A surround sound signal generating apparatus for generating a multi-channel surround sound signal for reproducing a sound signal in a selected sound field space, assuming a speaker arranged around a listener. Coefficient storing means for storing coefficients of frequency-to-level characteristics obtained by performing an FFT operation on impulse response characteristics for each direction of a sound field in a plurality of acoustic environments; and storing sound signal data of an incoming digital audio signal. F
FT operation, multiplying the data based on the operation result by the coefficient selected from the coefficient storage means, and performing an inverse FFT transform of the multiplied data, at least an operation processing means. Surround sound signal generator. 2. A surround sound signal generating apparatus for generating a multi-channel surround sound signal for reproducing a sound signal in a selected sound field space, assuming a speaker arranged around a listener. A coefficient storing means for dividing a plurality of impulse response characteristics for each direction of a sound field in a plurality of acoustic environments, and storing these divided characteristics as coefficients of frequency versus level characteristics obtained by performing FFT operations, respectively; FFT of audio signal data of incoming digital audio signal
Multiplying each of the divided coefficients of the impulse response characteristics for each direction selected from the coefficient storage means by the FFT-calculated data based on the incoming signal corresponding to the time length of these coefficients; A surround sound signal generating apparatus comprising at least an arithmetic unit for performing an inverse FFT transform on the multiplied data and adding the inverse FFT signals.