JP3855490B2 - Impulse response collecting method, sound effect adding device, and recording medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an impulse response collecting method, a sound effect adding device, and a recording medium that are particularly suitable for the surround effect when a reverberant sound is added based on an impulse response of an existing device or space.
[0002]
[Prior art]
One of devices for adding sound effects to an audio signal is a reverberation sound adding device (reverbulator). This reverberation adding device is often used, for example, to add reverberation to an audio signal in a recording studio and to make the sound broader and deeper. By adding reverberation to the sound recorded in a studio, you can add effects that are actually played in the hall, and even more special effects.
[0003]
In the old days, reverberation was added by recording in a place where reverberation was actually obtained, such as in a hall, or using a vibration such as a steel plate to obtain a reverberation effect. It was done using a device such as an iron plate echo. In recent reverberation sound adding apparatuses, these effects are electrically realized. Furthermore, in recent years, with the development of digital signal processing technology, devices that synthesize reverberant sound digitally have become widespread.
[0004]
When adding reverberation sound by digital processing, for example, a cyclic digital filter is used. The input digital audio signal is circulated while being attenuated to generate reverberant sound. This is mixed with the original digital audio signal. In practice, an initial reflected sound is added at a position delayed for a predetermined period with respect to the original sound, and a reverberant sound is added after a predetermined period. The delay time of the reverberant sound relative to the original sound is called pre-delay. With the addition of reverberation time and sub-reverberation sound, fine level adjustments can be performed, creating a wide range of sounds.
[0005]
By the way, a reproduction method called a surround method in which a sound field from the rear is added to a sound field reproduction from the left and right sound sources from the front is becoming widespread. The surround system is used especially for reproducing the sound field of a movie theater. For example, left and right channels (F-L, F-R) arranged in front of the listener and left and right channels arranged in the rear of the listener. 4 channels (R-L, R-R) are provided, and 1 channel (C) is arranged in the center in front of the listener. The C channel is for reproducing a speech in a movie sound, for example. Furthermore, one channel for reproducing the ultra-low range is arranged at an arbitrary position. This 5 channel + super low frequency channel configuration is expressed as 5.1 channels as a whole because the ultra low frequency requires only 1/10 of the amount of information transmitted.
[0006]
In addition to movies, normal music performances, etc., recorded in a surround format can be played back in a corresponding format so that you can enjoy a more realistic performance. In the case of music performance, normally, four channels of FL, FR, RL, and RR are used, and the C channel is not used.
[0007]
[Problems to be solved by the invention]
When recording in the surround system, it is necessary to add a reverberant sound corresponding to a sound field formed by four channels. Conventionally, two reverberation adding devices with monaural input / stereo output are often used. The reverberation sound adding device is a digital device in order to obtain a higher quality reverberation.
[0008]
For example, when recording one sound source assumed to be on the stage, the reverberation corresponding to the FL and FR channels assumed on the stage side with respect to the listener is performed by one reverberation adding device. And the reverberation corresponding to the RL and RR channels assumed behind the listener is added to the listener by the other reverberation sound adding device.
[0009]
Since the reverberation differs depending on the position, these two reverberation sound adding devices are set differently. As described above, conventionally, when recording corresponding to the surround system, it is necessary to use two reverberation sound adding devices, and thus there is a problem that the operation is inconvenient.
[0010]
In addition, the reverberant sound in an actual hall or the like has a more complicated waveform due to various reflections and interference of the sound depending on the shape of the hall and the position of the sound source. However, as described above, in the method of filtering the original digital audio signal, there is a problem in that an artificial impression is unavoidably inevitably obtained because a simply attenuated waveform is obtained.
[0011]
On the other hand, for example, a sound source having a predetermined sound field needs to be recorded in stereo. In this case, a reverberation adding device corresponding to stereo input is used. However, the conventional reverberation sound adding device corresponding to stereo input artificially creates a stereo sound image, and thus has a problem that the sound image is unnatural.
[0012]
In order to obtain a more natural reverberation, it is possible to record in an actual hall. However, recording in the hall has a problem that it takes a lot of time to carry in equipment. Also, there was a problem that the hall could not be borrowed when you wanted to borrow. Furthermore, recording in the hall has a problem that it requires know-how to set the microphone. Furthermore, in order to reduce background noise, there is a problem that it is necessary to stop an air conditioning facility or the like.
[0013]
Accordingly, an object of the present invention is to provide an impulse response collecting method, a sound effect adding device, and a recording medium that can easily obtain high-quality reverberation sound corresponding to the surround system.
[0015]
[Means for Solving the Problems]
  FirstThe invention relates to a method for collecting impulse responses used in generating sound effects by convolving impulse responses.Based on the measurement signal,Two places corresponding to the stereo systemOne and the otherRespectivelySelectivelyThe sound of the generated spaceIn correspondence with stereo system2 spatially arrangedPiecesWith each of the measuring means,EachA method for collecting impulse responses, comprising: a measuring step for measuring, and a converting step for converting each measured sound into an impulse response.
[0017]
  SecondThe inventionA computer-readable and removable recording medium based on a measurement signal,Two places corresponding to the stereo systemOne and the otherRespectivelySelectivelyThe sound of the generated spaceIn correspondence with stereo system2 spatially arrangedPiecesWith each of the measuring means,EachMeasurementHowever, the generated sound of the space for one and the other of the two locations corresponding to the stereo system isConvert each measured sound into an impulse responseOf 4 obtainedA computer-readable recording medium in which impulse response data is recorded.
[0018]
  ThirdThe invention provides a sound effect adding device for adding a sound effect by convolving impulse response data with respect to an input digital audio signal.Based on the measurement signal, the acoustics in the space selectively generated in one and the other of the two locations corresponding to the stereo system are respectively measured by the two measuring means spatially arranged corresponding to the stereo system. In each of the two locations corresponding to the stereo system, the sound of the generated space is measured and converted to impulse responses for each of the four acoustic waves.From a recording medium on which impulse response data is recorded4Reproducing means for reproducing the impulse response data of2 channels corresponding to stereo systemAn input means for inputting a digital audio signal;For each of the two-channel digital audio signals input to the input means,Played by playback meansFourImpulse response dataOut ofDigital audio signalBased on the impulse response data of 2 corresponding to the channel of the measurement meansEach convolution4 to do onlyConvolution means andOutput means for mixing the outputs of the four convolution means corresponding to the two locations corresponding to the stereo system in which the spatial sound is generated, and outputting the result as a two-channel digital audio signal;A sound effect adding device characterized by comprising:
[0020]
  As mentioned above,FirstThe inventionBased on the measurement signal,Two places corresponding to the stereo systemOne and the otherRespectivelySelectivelyThe sound of the generated spaceIn correspondence with stereo system2 spatially arrangedPiecesWith each of the measuring means,EachSince each measured sound is converted into an impulse response, an impulse response corresponding to a stereo input and a stereo output can be collected.
[0022]
  Also, as mentioned above,SecondThe inventionA computer-readable and removable recording medium based on a measurement signal,Two places corresponding to the stereo systemOne and the otherRespectivelySelectivelyThe sound of the generated spaceIn correspondence with stereo system2 spatially arrangedPiecesWith each of the measuring means,EachMeasure andFor each of one and the other of the two locations corresponding to the stereo system, the generated sound of the space isObtained by converting each measured sound into an impulse responseFourSince the impulse response data is recorded on the recording medium, the impulse response corresponding to the stereo input and the stereo output in the real space can be stored and reproduced.
[0023]
  As described above, the third aspect of the present inventionBased on the measurement signal, the acoustics in the space selectively generated in one and the other of the two locations corresponding to the stereo system are respectively measured by the two measuring means spatially arranged corresponding to the stereo system. In each of the two locations corresponding to the stereo system, the sound of the generated space was measured and converted to an impulse response.From a recording medium with impulse response data recorded4The impulse response data of2 channels corresponding to stereo systemDigital audio signalFor each ofPlayed4Impulse response dataEach of which is based on two impulse response data corresponding to the measurement means corresponding to the channel of the digital audio signal.Convolution4 convolution means are used, and the outputs of the 4 convolution means are mixed by 2 corresponding to the two locations corresponding to the stereo system where the sound of the space is generated, and output as 2-channel digital audio signals DoAs a result, the impulse response in the real space can be reproduced.
[0024]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described. The sound effect adding device in this embodiment is a reverberation adding device that adds a reverberation sound to an original sound composed of an input digital audio signal, and is an impulse obtained by collecting reverberation such as an actual hall. The input digital audio signal is convolved with the response data to obtain a reverberation sound to be added.
[0025]
FIG. 1 shows a reverberant sound according to this embodiment compared with a reverberant sound of a conventional recursive filter. The reverberant sound according to the prior art shown in FIG. 1A is delayed for a predetermined time with respect to the original sound to generate an initial reflected sound, and further delayed for a predetermined time and added with a reverberant sound generated by a filter. The reverberant sound is attenuated by a simple attenuation curve. On the other hand, in this embodiment, since the reverberation sound is generated by the impulse response based on the actually recorded data, the acoustic characteristics in the actual hall or the like are reflected as shown in FIG. 1B. A reverberant sound that is not a simple decay curve is obtained. Thereby, a more natural and high-quality reverberation sound can be obtained.
[0026]
The present invention provides an impulse response collection method for obtaining such a natural reverberation sound. In particular, the present invention provides an impulse response collection method suitable for use in the surround system. FIG. 2 shows an example of the configuration of the impulse response collection device 97 according to the present invention. In this example, the impulse response of the iron plate echo device 92 is measured. The impulse response collection device 97 can be configured by a personal computer, for example. In this apparatus 97, a signal for impulse response measurement is generated and output to a measurement object, the measurement result is collected, and the measurement result is converted into impulse response data. The impulse response data is stored as a file, for example.
[0027]
The measurement signal generator 90 generates a TSP (time stretch pulse) signal for measuring the impulse response. The TSP signal is a kind of sweep signal, and an impulse signal can be obtained by dividing by a signal having an inverse characteristic. In order to measure the impulse response, it is more preferable to directly generate the impulse signal, but such a method is used because measurement is difficult. The TSP signal generated by the measurement signal generator 90 is converted into an analog signal via the D / A converter 91 and input to the iron plate echo device 92.
[0028]
The iron plate echo device 92 generates a reverberation sound by the input TSP signal. The reverberant sound is output as analog audio signals of L (left) and R (right) channels. These outputs are converted into digital audio signals for the L and R channels by the A / D converter 93, respectively. In the A / D converter 93, for example, sampling is performed at a sampling frequency of 48 kHz or 96 kHz and a quantization bit number of 24 bits. As for the output of the A / D converter 93, each of the L and R channels is input to the impulse response collection device 97. The input signal is stored in, for example, a hard disk device or a memory (not shown).
[0029]
The reverberation time is determined as the time from when the sound stops until the sound pressure level is attenuated by 60 dB. In this example, in the quantization bit number of 24 bits, 6 dB is allocated to 1 bit.
[0030]
The generation of the TSP signal by the measurement signal generator 90 is performed N times. The output signals for N times are synchronously added by the synchronous adder 94. For example, the synchronous addition is performed by aligning the output signals based on the generation timing of the TS signal. By synchronously adding N signals, only reproducible signals are added, and randomly generated noise components are not added, so that the S / N ratio can be improved. The S / N ratio is improved by (10 log N) dB. For example, the S / N ratio is improved by 12 dB when N = 16.
[0031]
The L and R channels of the synchronously added signal are supplied to the impulse response converter 95. The impulse response conversion unit 95 divides the supplied signal by a signal having the inverse characteristic of the TSP signal. As a result, the TSP signal is converted into an impulse signal, and the measurement result is converted into an impulse response based on the reverberant sound generated by the impulse signal. The impulse response data is a peak value obtained at intervals corresponding to the sampling frequency. The signal sampled by the A / D converter 93 with a quantization bit number of 24 bits has a quantization bit number of 32 bits after conversion.
[0032]
The L channel impulse response data 96L and the R channel impulse response data 96R output from the impulse response converter are recorded on a predetermined recording medium such as a CD-ROM or MO. The impulse response collection device 97 may be provided with an interface such as Ethernet and supplied to the outside via a network.
[0033]
FIG. 3 shows an example of collecting impulse responses in the hall. The hall 101 has a stage portion 101A and a passenger seat portion 101B. The sound source 102 is placed at a predetermined position of the stage unit 101A. The sound source 102 is, for example, a dodecahedron speaker in which speakers are provided in 12 different directions on a spherical surface. In the passenger seat 101B, microphones 103L and 103R corresponding to the L and R channels, respectively, are placed at predetermined positions.
[0034]
The TSP signal output from the impulse response collection device 97 is converted into an analog signal by the D / A converter 91, amplified by the amplifier 100, and reproduced as sound by the sound source 102. This reproduced sound is recorded by the microphones 103L and 103R. The outputs of the microphones 103L and 103R are sampled by the A / D converter 93 at a predetermined sampling frequency and the number of quantization bits, respectively, and converted into L and R channel digital audio signals and supplied to the impulse response collecting device 97. The processing in the impulse response collection device 97 is exactly the same as the processing in the iron plate echo device 92 described above.
[0035]
In this case, impulse responses are collected by changing the position of the sound source 102 in various ways. The speakers used as the sound source 102 are also collected by changing the brands and the like. Similarly, the microphones 103L and 103R are recorded with various positions and brands. Thus, a plurality of data is collected in one hole 101. These can be selected as variations of reverberation sound, for example, when reverberation sound is added.
[0036]
On the other hand, the impulse response data 96L and 96R obtained by the impulse response converter 95 can be processed. FIG. 4 schematically shows the flow of processing when processing the impulse response data. The impulse response data 110 is processed 111. FIG. 5 shows an example of the processing process 111. As shown in FIG. 5A as an example, the data includes a system delay due to sound propagation (portion “A” in the figure). In the processing 111, the value of the system delay portion A is fixed at [0], and the noise in this portion is removed.
[0037]
In the second half of the data, fade-out processing is performed to converge the end of the data to [0]. By this fade-out processing, noise in the signal portion of the minute level in the latter half is also removed. 5B and 5C show an example of this fade-out process.
[0038]
FIG. 5B is an example in which fade-out processing is performed based on an exponential function of attenuation. For example, the original impulse response is h (n) and the fade-out function is F₀(N). n represents a point of impulse response data. The point of the impulse response data and the point of the sampling point of the digital audio signal correspond to each other. At this time, F₀In (n), if n ≦ 0, F₀(N) = 1. On the other hand, if n> 0, F₀(N) is an exponential function of attenuation as shown in FIG. 5B.
[0039]
The output data x (n) is expressed by the following equation (1):
x (n) = h (n) · F₀(Na) (1)
It becomes. The value a represents the position of the direct sound in the original impulse response by the number of samples. In this way, the fade-out is performed behind the direct sound position. This is because when the fade-out is started at the same position as the direct sound, that is, when n = 0, the level of the direct sound itself also decreases.
[0040]
The fade-out function is not limited to the exponential function of attenuation. For example, as shown in FIG. 5C, a linear attenuation characteristic may be used.
[0041]
Further, the number of points of the impulse response data can be adjusted by fading out so as to match the processing capability of the reverberation adding apparatus that actually adds the reverberation sound to the audio signal using this data. That is, the number of points of the impulse response data is described as a predetermined value, for example, 256 k points (262, 144 points: 256 k points with fractions omitted).ⁿ(For example, as shown in FIG. 4A, the fade-out starts at the point of 128k points and the data becomes [0] at the point of 256k points.) To do.
[0042]
As the processing process 111, in addition to the above, level adjustment and the like are also performed. The processed impulse response data is recorded, for example, on the CD-ROM 45 as the FIR filter coefficient 112 when convolved with the FIR filter.
[0043]
FIG. 6 schematically shows an example of the configuration of a reverberation adding apparatus that performs convolution using the impulse response data created in this way. A digital audio signal to which reverberant sound is to be added is input from the input terminal 120. The input data is supplied to the multiplier 126 and also delayed by the pre-delay 121 and given a pre-delay. The output of the pre-delay 121 is supplied to the convolution processing unit 122.
[0044]
The convolution processing unit 122 includes FIR filters (filter 122L and filter 122R) for the L and R channels, respectively. Impulse response data 96L and 97R created by the impulse response collection device 97 described above are supplied from terminals 123L and 123R as FIR filter coefficients of the corresponding channel. These impulse response data 96L and 96R are obtained by being read from, for example, a CD-ROM (not shown).
[0045]
The filters 122L and 122R convolve the input digital audio signal with the impulse response data 96L and 97R. As a result of this convolution, reverberant sound based on the impulse response data 96L and 96R is generated. Outputs of the filters 122L and 122R are supplied to multipliers 124L and 124R, respectively.
[0046]
The multipliers 124L and 124R, the above-described multiplier 126, and the adders 128L and 128R constitute a mixer of the original sound (dry component) and the reverberant sound (wet component). The multiplier 126 and the multipliers 124L and 124R adjust the input digital audio signal and the output of the convolution processing unit 122 according to the ratio between the original sound and the reverberation sound supplied to the terminals 127 and 125, respectively, and adders 128L and 128R. Thus, these signals are added, and the output of the L channel is derived to the output terminal 129L, and the output of the R channel is derived to the output terminal 129R.
[0047]
Furthermore, the present invention provides an impulse response collection method corresponding to the surround system. FIG. 7 shows an example of collecting impulse responses so as to correspond to the surround system. Impulse responses are collected in the hall 101 having the stage portion 101A and the passenger seat portion 101B, as described above. A sound source 102 made of, for example, a dodecahedron speaker is placed at a predetermined position of the stage unit 101A.
[0048]
The microphones 103FL, 103FR, 103RL, and 103RR are set at predetermined positions on the passenger seat 101B corresponding to the surround system in the passenger seat 101B. In other words, the microphones 103FL and 103FR correspond to the L and R channels (FL and FR channels) on the front side of the passenger seat 101B, respectively. The microphones 103RL and 103RR correspond to the rear L and R channels (RL and RR channels) of the passenger seat 101B, respectively.
[0049]
As an example, the positions of these microphones 103FL, 103FR, 103RL, and 103RR are sufficient to obtain a surround effect for the listener, assuming that the listener is seated at the optimum position of the passenger seat 101B with respect to the stage 101A. Position.
[0050]
The outputs of the microphones 103FL, 103FR, 103RL, and 103RR are supplied to an A / D converter 93 '. The A / D converter 93 'is an extension of the above-described A / D converter 93 so that signal processing for four channels is possible. In the A / D converter 93 ′, sampling is performed at a predetermined sampling frequency and the number of quantization bits, and FL (front left channel), FR (front right channel), RL (rear left channel), and R -R (Rear Right Channel) digital audio signal is supplied to the impulse response acquisition device 97 '.
[0051]
The impulse response collection device 97 'is an extension of the above-described impulse response collection device 97 so that signal processing for four channels can be performed. In the impulse response collection device '97, the same processing as that in the above-described iron plate echo device 92 is performed corresponding to the four channels FL, FR, RL, and RR, and the amount corresponding to the four channels. Impulse response data 96F-L, 96F-R, 96R-L and 96R-R are obtained (not shown).
[0052]
Even in this case, as described above, impulse responses are collected by changing the position of the sound source 102 in various ways. The speakers used as the sound source 102 are also collected by changing the brands and the like. Similarly, four microphones 103FL, 103FR, 103RL, and 103RR are also recorded by changing their positions and brands. Thus, a plurality of data is collected in one hole 101. These can be selected as variations of reverberation sound, for example, when reverberation sound is added.
[0053]
Also in this case, the impulse response data can be processed in the same manner as described above. The processed impulse response data is recorded on the CD-ROM 45 in the same manner.
[0054]
FIG. 8 shows an example of the configuration of a reverberation adding apparatus that adds reverberation sound corresponding to the surround system to input data based on the impulse responses collected based on the method shown in FIG. A digital audio signal to which reverberant sound is to be added is input from the input terminal 130. Input data is supplied to multiplier multipliers 136FL, 136FR, 136RL, and 136RR to obtain dry components of the FL, FR, RL, and RR channels, respectively, and FL, F In order to obtain wet components of the -R, RL, and RR channels, they are supplied to the pre-delays 131FL, 131FR, 131RL, and 131RR.
[0055]
In this example, each of the FL, FR, RL, and RR channels is independently processed in the same manner. Hereinafter, the processing for these four channels will be described by using the FL channel as a representative. The input data of the FL channel is delayed by the pre-delay 131FL and given a pre-delay. The output of the pre-delay 313FL is supplied to an FIR filter 132FL that performs convolution processing of impulse response data.
[0056]
The FIR filter 132FL is supplied with the F-L channel impulse response data 96F-L created by the above-described impulse response collection device 97 'from the terminal 133FL as a filter coefficient. These impulse response data 96F-L are obtained by being read from, for example, the CD-ROM 45 (not shown).
[0057]
In the FIR filter 132F-L, the input digital audio signal is convolved with the impulse response data 96F-L. As a result of this convolution, a reverberant sound based on the impulse response data 96F-L is generated. The output of the FIR filter 132F-L is supplied to the multiplier 134.
[0058]
The multiplier 134FL, the above-described multiplier 136FL, and the adder 138FL constitute a mixer for the original sound (dry component) and reverberation sound (wet component) of the FL channel. The multiplier 136FL and the multiplier 134FL adjust the input digital audio signal directly input from the input terminal 130 and the output of the FIR filter FL132 in accordance with the ratio between the original sound and the reverberation sound supplied to the terminals 137 and 135, respectively. These signals are added by the adder 138FL. The added output is led to the output terminal 139FL as the output of the FL channel.
[0059]
The other three channels are processed in the same manner as the FL channel. For example, the input data of the F-R channel is convolved with the impulse response data 96FR of the corresponding channel by the filter 132FR, and the ratio of the dry component to the wet component is obtained by the mixer including the multipliers 137FR and 136FR and the adder 138FR. Is output to the output terminal 139FR as the output of the FR channel.
[0060]
In this way, reverberant sound based on the impulse response data of the corresponding channel is added to each of the FL, FR, RL, and RR channels. The listener can reproduce the sound of the hall 101 in a surround manner. For example, the listener can enjoy the effect as if he was in the passenger seat 101B of the hall 101 by listening at a predetermined position during reproduction.
[0061]
FIG. 9 shows an example of collecting impulse responses so as to correspond to stereo inputs. In this case, sound sources 102L and 102R, for example, dodecahedron speakers, corresponding to the L and R channels, respectively, are placed on the stage unit 101A. The sound source 102L is set at a predetermined position on the left side from the passenger seat 101B, and the sound source 102R is set at a predetermined position on the right side. On the other hand, two microphones 103F-L and 103F-R respectively corresponding to the L and R channels are placed at predetermined positions in the passenger seat portion 101B, as in the example shown in FIG.
[0062]
The TSP signal from the amplifier 100 is supplied to the sound source selected by the selection unit 104 among the sound sources 102L and 102R. The selection unit 104 may be a switch, or the connection may be switched directly by detaching a cable.
[0063]
The impulse response is collected for each of the L and R channels. For example, the sound source 102 </ b> L is selected by the selection unit 104. The TSP signal is reproduced by the sound source 102L, and the reproduced sound is recorded by the microphones 103F-L and 103F-R. The outputs of the microphones 103F-L and 103F-R are supplied to the impulse response collecting device 97 via the A / D converter 93, subjected to impulse response conversion processing, and impulse responses of the L and R channels when the sound source is the L channel. Data 96L / F-L and 96L / F-R are obtained.
[0064]
Impulse responses are similarly collected for the sound source 102R, and TSP signals reproduced by the sound source 102R as selected by the selection unit 104 are recorded by the microphones 103F-L and 103F-R. Then, the impulse response data 96R / F-L and 96R / F-R of the L and R channels when the sound source is the R channel are obtained by the processing in the impulse response collection device 97.
[0065]
The impulse response data 96L / F-L, 96L / F-R, 96R / F-L, and 96R / F-R thus obtained are stored in the CD-ROM 45 after being subjected to predetermined processing, for example. To be recorded.
[0066]
FIG. 10 shows an example of the configuration of a reverberation adding apparatus that adds reverberation sound in stereo to data input in stereo based on impulse responses collected based on the method shown in FIG. Input data of the L channel is supplied from the input terminal 140L. Also, R channel input data is supplied from the input terminal 140R.
[0067]
First, the L channel processing will be described. The input data supplied from the input terminal 140L is supplied to a multiplier 146L that adjusts the ratio of the dry component, and is also supplied to the pre-delays 141LL and 141LR, and is given a pre-delay independently. The outputs of the pre-delays 141LL and 141LR are supplied to filters 142LL and 142LR that perform convolution of impulse responses, respectively.
[0068]
The filter 142LL performs a convolution process on the data supplied from the pre-delay 141LL based on the impulse response data 96L / F-L supplied from the terminal 143LL. The output of the filter 142LL is supplied to a multiplier 144LL that adjusts the ratio of the wet component.
[0069]
Similarly, a convolution process is performed on the output of the pre-delay 141LR by the filter 142LR using the impulse response data L / F-R supplied from the terminal 143LR. The output of the filter 142LR is supplied to a multiplier 144LR that adjusts the ratio of the wet component.
[0070]
Next, R channel processing will be described. The input data supplied from the input terminal 140R is supplied to a multiplier 146R that adjusts the ratio of the dry component, and is also supplied to predelays 141RL and 141RR, which are each given a predelay independently. The outputs of the pre-delays 141RL and 141RR are supplied to filters 142RL and 142RR that perform convolution of impulse responses, respectively.
[0071]
The filter 142RL performs a convolution process on the data supplied from the pre-delay 141RL based on the impulse response data 96R / F-L supplied from the terminal 143RL. The output of the filter 142RL is supplied to a multiplier 144RL that adjusts the ratio of the wet component.
[0072]
Similarly, a convolution process is performed on the output of the pre-delay 141RR by the filter 142RR using the impulse response data R / F-R supplied from the terminal 143RR. The output of the filter 142RR is supplied to a multiplier 144RR that adjusts the ratio of the wet component.
[0073]
Multipliers 144LL, 144LR, and 146L adjust the ratio of the dry component and the wet component with respect to the input of the L channel. The ratio of the dry component of the L channel is given from the terminal 147L to the multiplier 146L. Further, the ratio of the wet component is independently given from each of the terminals 145LL and 145LR to each of the multipliers 144LL and 144LR. The outputs of the multipliers 146L and 144LL are supplied to the adder 148L. The adder 148L is further supplied with the output of a later-described multiplier 144RL. These three outputs are added by the adder 148L, and the addition output is derived as an output of the L channel to the output terminal 149L.
[0074]
On the other hand, multipliers 144RR, 144RL, and 146R adjust the ratio of the dry component and the wet component with respect to the input of the R channel. The ratio of the dry component of the R channel is given from the terminal 147R to the multiplier 146R. Further, the ratio of the wet component is independently given from the terminals 145RR and 145RL to the multipliers 144RR and 144RL, respectively. The outputs of the multipliers 146R and 144RR are supplied to the adder 148R. The output of the above-described multiplier 144LR is further supplied to the adder 148R. These three outputs are added by the adder 148R, and the addition output is led to the output terminal 149R as the output of the R channel.
[0075]
In this way, in the case of stereo input support, L and R channel reverberation sounds from the sound sources 102L and 102R are added to the L and R channel input data, respectively. In the output, the reverberant sound is mixed for each channel. For each of the input L and R channels, a reverberation sound is added by convolving an impulse response actually collected in stereo for each L and R channel, so that a natural stereo sound image can be obtained.
[0076]
In addition, as indicated by a dotted line in FIG. 9, two microphones 103R-L and 103R-R may be further provided on the rear side. By doing so, it is possible to perform measurement corresponding to the surround system in addition to the stereo system. In this method, impulse responses from four microphones 103F-L, 103F-R, 103R-L, and 103R-R are collected for each of the L and R channel sound sources, and a total of 8 types of impulse response data are collected. Is obtained. Therefore, at the time of reproduction, another set of the configuration shown in FIG. 10 needs to be prepared. Similarly, although not shown, a center microphone may be added between the microphones 103F-L and 103F-R.
[0077]
FIG. 11 shows an example of the configuration of this reverberation adding apparatus more specifically. In the reverberation adding apparatus 1, digital audio signals for two channels (1ch / 2ch) are input from a digital audio input terminal 10 based on the AES / EBU (Audio Engineering Society / European Broadcasting Union) standard. The digital audio signal supplied from the input terminal 10 is supplied to the input switcher 12 via the digital input unit 11.
[0078]
The input digital audio signal has, for example, a sampling frequency of 48 kHz and a quantization bit number of 24 bits. In addition, by mounting the option board 50 described later on the apparatus 1, the sampling frequency that can be handled can be doubled to 96 kHz. Further, the present invention is not limited to these examples, and for example, a digital audio signal having a sampling frequency of 44.1 kHz can be supported. In this case, when the option board 50 is mounted, a signal with a sampling frequency of 88.2 kHz can be handled.
[0079]
When an analog audio signal is input to the reverberation adding device 1, analog audio input terminals 13L and 13R are used. Each of the audio signals of the L (left) and R (right) channels is input from the corresponding side of the input terminals 13L and 13R, and the A / D converter 14 has a quantization frequency of 24 bits at a sampling frequency of 48 kHz, for example. Is sampled and converted to a digital audio signal. The output of the A / D converter 14 is supplied to the input switcher 12.
[0080]
The input switcher 12 switches the system of the input audio signal by the control of the controller 40 described later or a manual changeover switch. The output of the input switcher 12 is supplied to a DSP (Digital Signal Processor) 30 through a path 31.
[0081]
The DSP 30 includes a DRAM (Dynamic Random Access Memory), and performs various controls of input / output digital audio signals based on a program supplied from a controller 40 described later. The DSP 30 supplies the supplied digital audio signal to the DSPs 32 </ b> A to 32 </ b> K for performing impulse response convolution based on a predetermined program. Further, the DSP 30 generates an initial reflected sound based on the input signal. Further, the impulse response convolution calculation result is supplied to the DSP 30 from the DSP 34 described later.
[0082]
The DSPs 32 </ b> A to 32 </ b> K cut out the digital audio signals supplied from the DSP 30 into blocks of a predetermined size, and perform a convolution operation based on impulse response data supplied in advance. The DSPs 32A to 32K each have a DRAM having a capacity corresponding to the number of samples to be processed. In this example, each of the DSPs 32A to 32H has two DRAMs, two DSPs 32I, four DSPs 32J and 32K, and a DRAM having a capacity of 16 Mbits.
[0083]
The result of convolution calculation of the impulse response for each block performed by the DSPs 32 </ b> A to 32 </ b> K is added by the adder 33 and supplied to the DSP 30 via the DSP 34. In the DSP 34, an overflow of the addition result is detected, and for example, the data causing the overflow is fixed to a predetermined value.
[0084]
The DSP 30 adds reverberant sound to the input digital audio signal by mixing the input digital audio signal, the above-mentioned initial reflected sound, and the result of convolution calculation of the impulse response supplied via the DSP 34. Output. The output 35 of the DSP 30 is supplied to the output switcher 18.
[0085]
The formed reverberant sound and the unprocessed input digital audio signal are also referred to as “wet component” and “dry component”, respectively. In the DSP 30, the mixing ratio of these wet components and dry components can be freely changed for each of the L and R channels. At the same time, the DSP 30 also adjusts the level of the output signal.
[0086]
The DSP 30 is supplied with a clock FS or 2FS having a frequency corresponding to the sampling frequency of the digital audio signal to be handled. Signal processing in the DSP 30 is performed based on this clock.
[0087]
The output switcher 18 switches the output signal system by the control of the controller 40 described later or a manual changeover switch. The output can be output as digital and analog audio signals. A digital audio signal for two channels is derived from the output switcher 18 through the digital output unit 19 to the output terminal 20 according to the AES / EBU standard. The digital audio signal output from the output switcher 18 is converted into analog audio signals of L and R channels by the D / A converter 21. The L and R channel analog audio signals are led to analog output terminals 22L and 22R, respectively.
[0088]
In this example, each of the input terminal 10, the input terminals 13L and 13R, the output terminal 20, and the output terminals 22L and 22R has a cannon type having three signal lines of hot, cold, and independent earth lines. It is used.
[0089]
The output switcher 18 can also be selected so as to bypass reverberation adding processing in the apparatus 1 for the input audio signal. When the bypass is selected, the input digital audio signal is directly supplied from the input switcher 12 through the bypass path 17 to the output switcher 18.
[0090]
On the other hand, the entire reverberation adding apparatus 1 is controlled by the controller 40. The controller 40 includes, for example, a CPU (Central Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), a predetermined input / output interface, and the like. In the ROM, for example, an initial program for starting the system and a serial number are stored in advance. The RAM is a work memory for the CPU to operate, and a program is loaded from the outside, for example.
[0091]
The controller 40 is connected to the bus 41 by, for example, 8-bit parallel. The bus 41 is connected to the above-described DSPs 30, 32A to 32H, 34, respectively. Communication is performed between the controller 40 and the DSPs 30, 32 </ b> A to 32 </ b> H, 34 via the bus 41. Through this communication, a program is supplied from the controller 40 to each of the DSPs 30, 32A to 32H, 34, and data and commands are exchanged between the controller 40 and the DSPs 30, 32A to 32H, 34. Is called.
[0092]
Further, as described above, the input switcher 12 and the output switcher 18 are connected to, for example, the bus 41 (not shown) and are controlled by the controller 40.
[0093]
For example, a display device 42 composed of a full-dot LCD (Liquid Crystal Display) is connected to the controller 40. Based on the display data generated by the controller 40, a predetermined display is performed on the display device 42.
[0094]
Although not shown, the input unit 43 includes a plurality of input means, for example, a rotary encoder configured to input data corresponding to a rotation angle, and a plurality of push switches. By operating these input means, corresponding control signals are supplied from the input unit 43 to the controller 40. Based on this control signal, a predetermined program, parameters, and the like are supplied from the controller 40 to the DSPs 30, 32A to 32H, 34.
[0095]
The reverberation adding apparatus 1 is provided with a CD-ROM (Compact Disc-ROM) drive 44. A CD-ROM 45 is inserted into the CD-ROM drive 44, and data and programs are read from the CD-ROM 45. The read data and program are supplied from the CD-ROM drive 44 to the controller 40.
[0096]
For example, impulse response data is recorded on the CD-ROM 45. The impulse response data is read from the CD-ROM 45 and supplied to the controller 40. Then, this data is supplied from the controller 40 to each of the DSPs 32A to 32K. The DSPs 32 </ b> A to 32 </ b> K perform a convolution operation of the impulse response based on the supplied impulse response data.
[0097]
By recording a large number of impulse response data collected in various environments on the CD-ROM 45, a reverberation effect similar to the environment corresponding to the impulse response to be used can be obtained. Also, a plurality of impulse response data can be used in combination. It is possible to create a space that does not actually exist. Further, the impulse response data can be processed by the reverberation adding device 1. For example, reverberation time is adjusted by processing the read impulse response data and performing fade-out processing.
[0098]
As another example, data obtained by converting impulse response data into frequency element data by Fourier transform may be recorded in the CD-ROM 45. The processing in the reverberation adding apparatus 1 can be reduced.
[0099]
Further, the CD-ROM 45 also stores display data used for display on the display unit 42 described above.
[0100]
The reverberation adding apparatus 1 includes MIDI (Musical Instrument Digital Interface) as an external interface. The MIDI signal supplied from the MIDI input terminal 46 is supplied to the controller 40. Based on the supplied MIDI signal, predetermined functions of the apparatus 1 can be controlled. Further, the controller 40 can generate and output a MIDI signal. The MIDI signal supplied from the MIDI input terminal 46 can be processed and output. The MIDI signal output from the controller 40 is supplied from the MIDI output terminal 47 to an external device. The MIDI through terminal 48 outputs the MIDI signal supplied from the MIDI input terminal 46 as it is.
[0101]
The reverberation adding apparatus 1 can be expanded in function by installing the option board 50. As an example of the function expansion, a digital audio signal with a sampling frequency of 48 kHz can be further handled with two systems, that is, signals for a total of four channels. As a result, for example, the above-mentioned surround reverberation sound addition and stereo input / stereo output reverberation sound addition can be performed by one reverberation addition device 1.
[0102]
As another example of the function expansion, when a digital audio signal for two channels (1ch / 2ch) is handled, a signal whose sampling frequency is 96 kHz, which is twice as high, can be handled.
[0103]
Digital audio signals for two channels (3ch / 4ch) are input from the terminal 15 via the option board 50. This digital audio signal is supplied to the input switcher 12 via the digital input unit 16. Further, the digital audio signals for two channels corresponding to the processing in the option board 50 output from the output switcher 18 are led to the terminal 24 via the digital output unit 23. This digital audio signal is output from the terminal 24 to the outside via the option board 50.
[0104]
The option board 50 and the device 1 are connected to each other through terminals 51 to 56 and terminals 15 and 24. FIG. 12 shows an example of the configuration of the option board 50. The option board 50 is configured to extend and execute the impulse response convolution operation by the DSPs 32A to 32K and the adder 33 described above. Therefore, the option board 50 is provided with DSPs 32L, 32M, and DSPs 60A-L similar to the above-described DSPs 32A-32K, and an adder 61 and a DSP 62 corresponding to the DSP 34 described above.
[0105]
The bus 41 ′ on the board 50 is connected to the bus 41 of the device 1 through the terminal 56. The DSPs 32L, 32M, and DSPs 60A to 60L on the board 50 can communicate with the controller 40 via the bus 41 '.
[0106]
The DSPs 32L and 32M have eight 16M-bit DRAMs, and perform a convolution operation together with the DSPs 32A to K described above. An input digital audio signal is output from the DSP 30 and supplied to the DSPs 32L and 32M via the terminal 53, respectively. The convolution calculation results by the DSPs 32L and 32M are supplied to the adder 33 via the terminals 54 and 55, respectively, and added together with the calculation results of the other DSPs 32A to 32K.
[0107]
On the other hand, the DSPs 60A to 60M perform processing in parallel with, for example, the above-described DSPs 32A to 32M. An input digital audio signal is output from the DSP 30 and distributed to the DSPs 60 </ b> A to 60 </ b> M via the terminal 51.
[0108]
For example, when processing for 4 channels from 1ch to 4ch is performed by installing the option board 50, the 1ch and 2ch convolution operations are performed by the DSPs 32A to 32M, and the 3ch and 4ch convolutions are performed by the DSPs 60A to 60M. An operation is performed. When handling a digital audio signal with a sampling frequency of 96 kHz, for example, DSPs supplied with blocks having the same number of samples, that is, DSPs 32A and 60A, DSPs 32B and 60B,..., DSPs 32M and 60M, respectively. By performing the convolution operation in parallel, processing at double speed can be handled.
[0109]
The convolution calculation results in the DSPs 60A to 60M are respectively supplied to the adder 61 and added. The addition result is supplied to the DSP 62, subjected to overflow processing in the same manner as the DSP 34 described above, and supplied to the DSP 30 via the terminal 52. In the DSP 30, the ratio of the dry component and the wet component is adjusted as necessary, and the mixing ratio with the signals of other channels is adjusted and supplied to the output switcher 18.
[0110]
The option board 50 is provided with an input terminal 63 and an output terminal 64 for digital audio signals based on the AES / EBU standard. Signals for two channels (3ch / 4ch) are input to the input terminal 63, and the input signals are supplied to the input switcher 12 via the terminal 15. Similarly, output signals for two channels (3ch / 4ch) output from the output switcher 18 are supplied to the board 50 via the terminal 24 and led to the output terminal 64. In this example, the terminals 63 and 64 are Canon type.
[0111]
FIG. 13 shows an example of the front panel 200 of the reverberation adding apparatus 1. Mounting holes are provided at the four corners of the front panel 200 so that the device 1 can be mounted on a rack. A power switch 201 is provided on the left side of the panel 200, and a CD-ROM insertion part 202 for attaching the CD-ROM 45 to the CD-ROM drive 44 is provided below the power switch 201. By operating the switch 205, it is possible to insert the CD-ROM 45 into the CD-ROM insertion section 202 and to remove the CD-ROM 45 from the insertion section 202.
[0112]
A display unit 203 is provided at a substantially central portion of the panel 200. The display unit 203 corresponds to the LCD 42 described above. A rotary encoder 204 is provided on the right side of the display unit 203. In addition, function keys 206, 207, 208, and 209 are provided below the display unit 203. The rotary encoder 204 and function keys 206 to 209 can be used to select functions of the apparatus 1 and input data.
[0113]
The display unit 203 performs various displays according to the selected function. In this example, parameter display is performed when a predetermined reverberation type is selected, and the parameters specified for the selected reverberation sound are sensed in the display area 210 in the display unit 203. In addition to the display, the display area 211 displays parameter names and parameter values.
[0114]
The display in the display area 211 corresponds to each of the function switches 206 to 209 arranged at the lower part of the display unit 203. For example, by pressing any one of the function keys 206 to 209, the parameter displayed immediately above the pressed key is selected. When the rotary encoder 204 is rotated, the parameter is changed. In addition, for example, another page can be displayed on the display unit 203 by a predetermined operation. On different pages, different parameter values can be changed.
[0115]
On the other hand, in this embodiment, a ripple corresponding to the currently set parameter value is displayed in the display area 210, and the effect (sound spread) of the reverberant sound by the parameter value can be grasped sensuously. Has been. 10 and 11 show examples of display in the display area 210. FIG. As the reverberation time is changed from a short value to a long value, the wave number of the ripple is increased as shown in FIGS. 14A to 14H and FIGS. 15A to 15H.
[0116]
In this example, the ripple has a 16-step display corresponding stepwise to values from the minimum value to the maximum value of the reverberation time. This 16-step display is relative to the reverberation time. Display data for displaying ripples is stored in the CD-ROM 45. For example, when the apparatus 1 is activated, it is read from the CD-ROM 45 in advance and stored in the RAM of the controller 40. Not limited to this, it may be stored in advance in the ROM of the controller 40. When the parameters of the reverberation time are determined, the display of the ripples is retained at that time.
[0117]
By performing such display, a visual impression can be given to the user. The user can grasp the effect of reverberation sensuously. That is, the user can visually grasp the spread of the reverberant sound by the ripples.
[0118]
In this example, the ripples are displayed so as to spread from the lower left to the upper right of the display area 210, but this is not limited to this example. FIG. 12 shows another example of ripple display on the display area 210. The center point of the ripple and the direction in which the ripple spreads can be arbitrarily set. For example, the left end can be the center of the ripple (FIG. 16A). In addition, the center of the display area 210 can be the center of the ripples (FIG. 16B). Furthermore, a ripple cross section may be displayed (FIG. 16C). Further, the shape of the ripples can be changed according to the type of the selected reverberation sound. Furthermore, in this example, the ripples are displayed in a fixed manner, but a plurality of pieces of display data are prepared for one stage parameter, and these are continuously switched to display an animation display. You can also
[0119]
Next, the impulse response convolution calculation performed by the DSPs 32A to 32M and the DSPs 60A to 60M will be described. Here, in order to avoid complication, the calculation performed only by the DSPs 32A to 32K without using the option board 50 will be described.
[0120]
FIG. 17 schematically shows processing in each of the DSPs 32A to 32K. The impulse response data is read from, for example, the CD-ROM 45 under the control of the controller 40, supplied in advance to the DSPs 32A to 32K, and stored in DRAMs provided in the DSPs 32A to 32K, respectively. Then, in each of the DSPs 32A to 32K, the impulse response data is divided at predetermined intervals on the time axis corresponding to the processing block size determined for each.
[0121]
Here, the DSPs 32 </ b> A to 32 </ b> K are represented as the DSP 32, and the unit of the impulse response processed by the DSP 32 is N. For example, in this example, N = 128 because the DSP 32A is designed to perform a convolution operation of impulse response data of 128 points. In the following description, one word corresponds to one sampling data of a digital audio signal. Therefore, one word has a time interval of (1 / sampling frequency) on the time axis, and the digital data has a quantization bit number (24 bits).
[0122]
The input data supplied to the DSP 32 is cut into block data consisting of N words. Therefore, the time for the first N words is spent inputting data. The input N word data is stored in the DRAM of the DSP 32. Then, the impulse response convolution operation is performed on the stored input data for N words in the time corresponding to the next N words. When all the calculations are completed, the calculation results for N words are output. Therefore, in the calculation of N words, a delay of 2N words occurs with respect to data input / output.
[0123]
FIG. 18 shows the process in the DSP 32 in more detail. In the DSP 32, a convolution calculation of an impulse response is performed using an overlap save method in a cyclic convolution which is a well-known technique.
[0124]
That is, as shown in FIG. 18, DFT (Discrete) is applied to the nth block 80B and the previous (n−1) th block 80A supplied every N words according to the time axis. (Fourier Transform) is performed to convert the data on the time axis into frequency element data 81 composed of a real part 81A of (N + 1) words and an imaginary part 81B of (N-1) words.
[0125]
On the other hand, the impulse response data 82 is DFT in advance with respect to N words of real data 82A and zero data 82B, respectively, and frequency element data comprising a real part 83A of (N + 1) words and an imaginary part 83B of (N-1) words. 83.
[0126]
The frequency elements corresponding to each other of the frequency element data 81 based on the input data and the frequency element data 83 based on the impulse response are multiplied with each other, and a filtering process (convolution) is performed to add equal frequency components to the multiplication result. As a result of this calculation, frequency element data 84 including a real part 84A of (N + 1) words and an imaginary part 84B of (N-1) words is obtained. The frequency element data 84 is subjected to IDFT, which is the inverse process of DFT, to obtain data 86 on the time axis composed of 2N words.
[0127]
As shown in the data 85, 86, and 87 in FIG. 18, IDFT results are obtained 2N words at intervals of N words. In each of the data 85, 86, and 87, the first half N words of data 85A, 86A, and 87A are discarded, and the (n-1) th block, the nth block, the (n + 1) th block, and so on. In this way, output data is obtained. The nth output data is delayed by 2 blocks with respect to the corresponding nth input data.
[0128]
A long reverberation time can be obtained by increasing the block size and performing a convolution calculation of more impulse response data in a single process. However, as described above, there is a delay of two blocks until the input block is output. Therefore, if one block is increased, the delay time until the reverberation processing component is output becomes longer. Not practical. Therefore, in this embodiment, processing for obtaining a desired reverberation time is performed in parallel for each of a plurality of blocks divided into a predetermined number of points (number of words).
[0129]
FIG. 19 and FIG. 20 show the convolution calculation processing divided into a plurality of blocks according to this embodiment. For example 2¹⁸Consider a case where a word (256k word) convolution operation is performed. In this case, the digital audio signal is convolved with impulse response data of 256k words (256k points). A reverberation time of approximately 5.3 sec is obtained when the sampling frequency is 48 kHz, and a reverberation time of approximately 5.9 sec is obtained when the sampling frequency is 44.1 kHz.
[0130]
As shown in an example in FIG. 19, the entire 256k word is divided into two parts, and the side located on the time axis among the two parts is further divided into two parts. In this way, the side positioned in front on the time axis is sequentially divided into two. Of the two divided parts, each of the rear side on the time axis is further divided into two to form two blocks of the same size.
[0131]
FIG. 20 shows an enlarged portion A of the first 8k words in FIG. This portion A is also divided into two in the same manner, but for the first 256 words, two 128-word blocks are formed, and impulse response convolution is performed for these two blocks. Therefore, the reverberation component is output after being delayed by the first 256 words. However, for example, when the sampling frequency is 48 kHz, this is a delay of only 5 msec, and there is no problem in terms of reverberation.
[0132]
Thus, the whole is 2¹⁸In this example of words (256k words), 2⁷Word (128 words), 2⁸Word (256 words), 2⁹Word (512 words), 2^TenWord (1k word), 2¹¹Word (2k words), 2¹²Word (4k words), 2¹³Word (8k words), 2¹⁴Word (16k words), 2¹⁵Word (32k words) and 2¹⁶2 words (64k words)ⁿTwo blocks each having a word size are formed.
[0133]
In the DSPs 32 </ b> A to 32 </ b> K, processing is performed for each group having the same block size. That is, as shown in FIGS. 19 and 20, the input data supplied to the DSPs 32A to 32K is 128 words for the DSP 32A, 256 words for the DSP 32B, 512 words for the DSP 32C, and 512 words for the DSP 32D. 1k word, 2k word with DSP32E, 4k word with DSP32F, 8k word with DSP32G, 16k word with DSP32H, 32k word with DSP32I, and 64k word with DSP32J and 32K.
[0134]
In each of the processes from 128 words to 32k words, the convolution process for two blocks having the same block size is performed in a time division manner by one DSP.
[0135]
That is, in each of the DSPs 32 </ b> A to 32 </ b> K, a convolution operation is performed on the extracted block data using the corresponding impulse response data. The latter half blocks of the set having the same block size are output after being delayed by one block after processing. As a result, in each of the DSPs 32A to 32K, two blocks of the same size are continuously output. By adding the outputs of the DSPs 32A to 32K by the adder 33, reverberation data 88 is generated.
[0136]
The data continuously supplied to each of the DSPs 32A to 32K is processed in the respective cycles of the DSPs 32A to 32K, and the results are added to thereby continuously supply the data. It is well known that reverberant sound can be added to.
[0137]
FIG. 21 shows an example of the configuration of a convolution filter 70 for performing a convolution operation in each of the DSPs 32A to 32K. The convolution filter 70 is realized based on a predetermined program supplied from the controller 40 to the DSPs 32A to 32K, for example. A digital audio signal is input from the terminal 71 and supplied to the DFT circuit 72. The digital audio signal is converted from data on the time axis into frequency element data by the DFT circuit 72. The output of the DFT circuit 72 is supplied to the multiplier 74 and also to the delay circuit 73.
[0138]
The delay circuit 73 has a delay of N words. That is, the DSPs 32A to 32K have N = 128, 256, 512, 1k, 2k, 4k, 8k, 16k, 32k, and 64k, respectively, and have corresponding delay amounts. The data delayed by the delay circuit 73 is supplied to the multiplier 76.
[0139]
The multiplier 74 is supplied with a filter coefficient A, which is DFT-impulse response data, from a terminal 75. The multiplier 74 multiplies the output of the DFT circuit 72 and the corresponding frequency element of the filter coefficient A. On the other hand, the multiplier 76 performs the same processing. That is, the filter coefficient B which is the DFT impulse response data is supplied from the terminal 77, and the output from the delay circuit 73 and the corresponding frequency element of the filter coefficient B are multiplied.
[0140]
The multiplication results of the multipliers 74 and 76 are added by an adder 78. The addition result is supplied to the IDFT circuit 79, and the frequency element data is converted into data on the time axis and output from the terminal 80.
[0141]
As described above, the convolution filter 70 performs a convolution operation using two blocks of data, that is, input data and N words, that is, input data delayed by one block, so that two words of data are obtained. Is output. As already described with reference to FIG. 14, the first half of the output data of two words is discarded.
[0142]
FIG. 22 shows processing of the convolution filter 70 based on the configuration of FIG. 21 described above corresponding to the time axis. The input data is shown on the left end side of FIG. 22, and the output data is shown on the right end side. FIG. 22 shows the passage of time from the upper side to the lower side as a whole. That is, although it has been shown that there are a plurality of filters 70, these indicate processing at different timings of one filter. In this manner, the result of DFT at the previous timing is delayed by the delay circuit 73 and used for the filter processing at the next timing. Therefore, output data delayed by two blocks with respect to input data is continuously output.
[0143]
FIG. 23 is a functional block diagram showing an outline of parallel processing of the DSPs 32A to 32K. Input data is supplied in parallel to each of the DSPs 32A to 32K. DSPs 32A-32K convolve points of N = 128, N = 256, N = 512, N = 1k, N = 2k, N = 4k, N = 8k, N = 16k, N = 32k and N = 64k, respectively. I do. The calculation result is delayed by 2N words from each of the DSPs 32 </ b> A to 32 </ b> K and supplied to the adder 22.
[0144]
For example, the input data supplied to the DSP 32A is cut out into blocks each consisting of N = 128 words, convolved with the cut out blocks, delayed by 2N words with respect to the input timing, and the operation result is output. The Then, the next block of N words is fetched and the same processing is repeated. The same processing is performed in each of the DSPs 32B to 32K.
[0145]
In the above description, the impulse response collection device 97 and the reverberation adding device 1 are described as separate devices, but this is not limited to this example. That is, the reverberation adding apparatus 1 is provided with a measurement signal generation unit 90 that generates a TSP signal, a synchronous addition unit 94, and an impulse response conversion unit 95. Needless to say, these can be constituted by a CPU and some peripheral components. It is also possible to use the DSP 30 or the DSP 34 that the reverberation adding apparatus 1 originally has. In this way, by providing the reverberation adding apparatus 1 with a function of collecting impulse responses, it is possible to obtain user-specific sound effects.
[0146]
In the above description, the impulse response convolution process is performed by hardware such as DSPs 32 </ b> A to 32 </ b> K. However, this is not limited to this example, and may be performed by software processing. Similarly, the processing of the DSPs 30 and 34 can be performed by software.
[0147]
Further, in the above description, the present invention is described as corresponding to the 5.1 channel or 4 channel surround system, but this is not limited to this example. By setting the microphone at a position corresponding to the reproduction system when collecting the impulse response, the present invention arranges, for example, two speakers for each of the front L and R channels, and two L and R channels at the rear. The present invention can also be applied to a system in which a book speaker is arranged.
[0148]
【The invention's effect】
As described above, according to one embodiment of the present invention, the impulse response actually obtained by the measurement in the real space is convolved with each other to generate the reverberation sound, so that natural and high-quality processing is performed. There is an effect that a result is obtained. That is, according to this embodiment, there is an effect that the pitch of the reverberant sound becomes equal to the pitch of the input sound.
[0149]
Further, according to one embodiment of the present invention, the impulse response is obtained by performing the measurement with the TSP signal a plurality of times and synchronously adding the obtained results for a plurality of times. There is an effect that it is possible to realize a higher S / N ratio than the reverberation sound obtained by an actual space.
[0150]
Furthermore, according to an embodiment of the present invention, by processing the impulse response data obtained from the real space, there is an effect that the reverberation time can be adjusted, which cannot be performed in the actual space.
[0151]
Furthermore, according to the present invention, since the output for four devices can be controlled by one device, there is an effect that the operation is easy.
[0152]
In addition, according to the present invention, there is an effect that natural sound image localization can be obtained even at the time of stereo input.
[0153]
Furthermore, according to the present invention, there is an effect that the sound effect of the microphone setting similar to the recording in the actual hall can be easily obtained.
[Brief description of the drawings]
FIG. 1 is a schematic diagram showing a reverberant sound according to an embodiment compared with a reverberant sound of a conventional recursive filter.
FIG. 2 is a block diagram showing an example of the configuration of an impulse response collection device according to the present invention.
FIG. 3 is a schematic diagram illustrating an example of collecting impulse responses at holes.
FIG. 4 is a schematic diagram illustrating an example of impulse response processing.
FIG. 5 is a schematic diagram illustrating an example of impulse response processing.
FIG. 6 is a block diagram schematically showing an example of the configuration of a reverberation adding apparatus that performs convolution using impulse response data.
FIG. 7 is a schematic diagram illustrating an example when impulse responses are collected so as to correspond to the surround system;
FIG. 8 is a block diagram illustrating an example of a configuration of a reverberation adding apparatus that adds reverberant sound to input data in accordance with the surround method.
FIG. 9 is a schematic diagram illustrating an example of collecting impulse responses so as to correspond to a stereo input;
FIG. 10 is a block diagram illustrating an example of a configuration of a reverberation adding apparatus that adds reverberant sound in stereo to data input in stereo.
FIG. 11 is a block diagram showing more specifically an example of the configuration of the reverberation adding apparatus.
FIG. 12 is a block diagram illustrating an example of a configuration of an option board of the reverberation adding device.
FIG. 13 is a schematic diagram illustrating an example of a front panel of a reverberation adding device.
FIG. 14 is a schematic diagram illustrating an example of ripples displayed in a display area.
FIG. 15 is a schematic diagram illustrating an example of ripples displayed in a display area.
FIG. 16 is a schematic diagram illustrating another example of ripples displayed in the display area.
FIG. 17 is a schematic diagram schematically illustrating processing in each DSP that performs a convolution operation;
FIG. 18 is a schematic diagram showing the process in each DSP in more detail.
FIG. 19 is a schematic diagram illustrating a convolution operation process divided into a plurality of blocks.
FIG. 20 is a schematic diagram illustrating a convolution operation process divided into a plurality of blocks.
FIG. 21 is a block diagram illustrating an example of a configuration of a convolution filter in each DSP.
FIG. 22 is a schematic diagram illustrating convolution filter processing corresponding to a time axis;
FIG. 23 is a schematic diagram illustrating an example in which different N words are processed in parallel.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Reverberation addition apparatus, 30 ... DSP, 32A-32M ... DSP, 33 ... Adder, 34 ... DSP, 40 ... Controller, 42 ... Display part by LCD, 43 ... Input unit, 44 ... CD-ROM drive, 45 ... CD-ROM, 50 ... Option board, 60A-60M ... DSP, 61 ... Adder, 62 ... DSP, 90... Measurement signal generator, 94... Synchronous adder, 95... Impulse response converter 95, 96L, 96R, 96F-L, 96F-R, 96R-L, 96R-R, 96L / F-L, 96L / F-R, 96R / F-L, 96R / F-R ... Impulse response data, 97 ... Impulse response collection device, 101 ... Hall, 102L, 102R ...・ Sound source, 103FL, 103 R, 103LR, 103RR ··· microphone, convolution 122 ... filter

Claims (6)

In the impulse response collection method used when generating sound effects by convolving the impulse response,
Based on the measurement signal, the sound of each selectively generated spatial in one and the other of the two positions corresponding to the stereo system, each of the two measuring means spatially arranged to correspond to the stereo system And each measuring step to measure,
And a step of converting each of the measured sounds into an impulse response. The impulse response collection method according to claim 1 ,
The measurement signal is generated a plurality of times, and further includes a step of synchronous addition for synchronously adding the measurement results of the measurement signal generated a plurality of times.
In the conversion step, the conversion to the impulse response is performed based on the addition result in the synchronous addition step. The impulse response collection method according to claim 1 ,
The impulse response collection method further comprising a machining step of machining the impulse response converted in the conversion step. A recording medium readable and removable by a computer,
Based on the measurement signal, the sound of each selectively generated spatial in one and the other of the two positions corresponding to the stereo system, each of the two measuring means spatially arranged to correspond to the stereo system In each of the two locations corresponding to the stereo system, the generated sound in the space is obtained by converting each of the measured sounds into an impulse response. A computer-readable recording medium in which the impulse response data is recorded. The recording medium according to claim 4 ,
A computer-readable recording medium, wherein conversion to the impulse response is performed using an addition result obtained by synchronously adding measurement results obtained by the measurement signal generated a plurality of times. In a sound effect adding device for adding a sound effect by convolving impulse response data with respect to an input digital audio signal,
Based on the measurement signal, the acoustics in the space selectively generated in one and the other of the two locations corresponding to the stereo system are respectively measured by the two measuring means spatially arranged corresponding to the stereo system. In each of the two locations corresponding to the stereo system, the generated sound in the space is obtained by converting each of the measured sounds into an impulse response. Reproducing means for reproducing the four impulse response data from a recording medium on which the impulse response data is recorded;
Input means for inputting a two-channel digital audio signal corresponding to a stereo system ;
Of the four impulse response data reproduced by the reproducing means for each of the two-channel digital audio signals inputted to the input means, the measuring means corresponds to the channel of the digital audio signal. means convolution of 4 for convolution themselves respectively based on the impulse response data,
Output means that mixes two outputs of the convolution means of 4 corresponding to two locations corresponding to the stereo system in which the sound of the space is generated, and outputs the mixture as a 2-channel digital audio signal; A sound effect adding device characterized by comprising:

Images