JPH0455896A - Signal processor - Google Patents

Abstract

PURPOSE:To form reverberation sounds of lower periodicity and long reverberation time with a limited number of steps by attenuating the delay signal obtd. by a signal delaying section by as much as a predetermined ratio and feeding the signal back to the input of a unidirectional signal processing section. CONSTITUTION:The signal delaying section 2 having the delay time larger than the total sum of the delay time of the input signals generated respectively by plural stages of delaying devices in the unidirectional signal processing section 1 consisting of noncyclic type filters, etc., is connected to the input of the unidirectional signal processing section 1 in the case of forming of the output signals having the prescribed reverberation characteristics, etc., by processing the input signal Si in one direction by means of the above-mentioned delaying devices. The delay signal Sd obtd. by the signal delaying section 2 is attenuated by as much as the predetermined ratio and is then fed back to the input of the unidirectional signal processing section 1. The reverberation sounds of the lower periodicity and the long reverberation time are formed in this way with the limited number of steps.

Description

[Detailed Description of the Invention] [Summary] A human signal is processed in one direction by a multi-stage delay device in a unidirectional signal processing unit consisting of an acyclic filter, etc., to produce an output signal having predetermined reverberation characteristics, etc. When generating, a signal delay section having a delay time larger than the sum of delay times of the input signals caused by each of the delay devices is connected to the input of the unidirectional signal processing section, and the delay obtained by the signal delay section is connected to the input of the unidirectional signal processing section. A signal processing device in which a signal is attenuated by a predetermined percentage and then returned to the human power of a unidirectional signal processing section, and the reverberation sound with little periodicity and long reverberation time is processed in a limited step. It becomes possible to generate by numbers.

[Industrial application field]

The present invention relates to a signal processing device for processing an input signal to generate an output signal having desired reverberation characteristics and the like.

For example, in audio systems, in order to give the sound a sense of presence, it is required to add appropriate reverberation sound (also called reflected sound) to the direct sound from various sources such as cassette tape recorders and record players. There are cases. The present invention refers to a signal processing device having a function of generating such reverberant sound.

[Conventional technology]

When generating a desired reverberant sound in an audio system or the like, a digital filter is usually used in which the amplitude, reverberation time, etc. of the reverberant sound can be easily set using software. More specifically, this digital filter is composed of a delay device, a multiplier, and an adder, and the coefficients of each of these elements are set to appropriate values by a program of a microcomputer or the like. If the analog audio signal output from the source is converted into a digital audio signal and then input to a digital filter, the audio input signal will be appropriately processed by the delay device etc. and will have the predetermined reverberation characteristics. An output signal is generated. Further, this output signal is D/A converted and then reproduced as reverberant sound by a speaker or the like.

Here, if you want to generate a relatively long reverberant sound with a small number of elements (number of program steps), use the digital
As a filter, generally a cyclic filter (Infinite Impu
lse Re5ponse Filter, hereinafter, II
There is a tendency for R filters to be used. This II
As shown in FIG. 5, the reverberant sound obtained by the R filter is generated periodically at fixed time intervals defined by the delay time T and is repeated infinitely.

Therefore, the reverberation time until the reverberation sound falls below a certain level can be set to an arbitrary length.

On the other hand, if you want to create irregularities in the time intervals between each reverberant sound and their amplitudes, you can use a type of non-standard method in which the input signal is processed while proceeding in one direction using multiple stages of delay devices and multipliers. Cyclic filter (Finite Impu
lse Re5ponse Filter, hereinafter FI
There is a tendency for R filters to be used. In this case, as shown in FIG. 6, the delay times T1 to TI in each of the multiple stages of delay devices and multipliers.

By setting the multiplication coefficients g, to g, to arbitrary values by a program, it is possible to generate reverberant sound without periodicity. However, in the FIR filter, unlike the IIR filter, the output signal is not fed back to the input, so the number of times reverberation occurs is determined by the number of stages of delay devices and the like. Since the number of program steps is usually limited, the number of stages of delay devices and the like is limited, and as a result, the length of the reverberation time is also limited.

[Problem to be solved by the invention]

As mentioned above, when generating the desired reverberant sound in an audio system, etc., conventionally, the gain can be obtained by processing the input signal by using either an IIR filter or a FIR filter, or by adding both in parallel. The output signal is played back by a speaker or the like.

In this case, the above-mentioned 11R filter has the advantage of being able to set a long reverberation time, but since periodicity (coloring) occurs in the reverberant sound, standing waves are likely to occur, and depending on the user's preference, a concert pole or A problem arises in that it becomes difficult to quickly set up the sound field of a live house or the like.

On the other hand, although FIR filters do not have the problem of periodicity in reverberant sound, in order to obtain the long reverberation time required in a specific sound field setting, the number of steps in the program becomes too large for practical use. The problem arises that it becomes difficult. For this reason, the above IIR filter and F
With any of the IR filters, it is difficult to generate reverberant sound that fully satisfies the user's requirements.

The present invention has been made in view of the above problems, and an object of the present invention is to provide a signal processing device that can generate reverberant sound with less periodicity and a longer reverberation time in a limited number of steps. It is something to do.

[Means to solve the problem]

FIG. 1 is a block diagram showing the basic configuration of the present invention. Here, a digital filter 10 for processing an input signal Si to generate a predetermined output signal S0 includes a unidirectional signal processing section 1 including a plurality of stages of delay devices. In this unidirectional signal processing section 1, the human signal S is processed while passing in one direction through the plurality of stages of delay devices. Furthermore, a signal delay section 2 is connected to the input of the unidirectional signal processing section 1. This signal delay section 2 has a delay time that is larger than the sum of the delay times of the input signal S1 caused by each of the delay devices. Further, the digital filter 10 receives the delayed signal S6 obtained by the signal delay section 2.
A signal attenuation section 3 is provided for attenuating the signal by a predetermined ratio and returning the unidirectional signal processing section 10 to human power.

[For production]

According to FIG. 1, reverberant sounds are sequentially generated at different time intervals using individually determined delay times in each of the multiple stages of delay devices in the unidirectional signal processing section 1, which is composed of an FIR filter, etc. ing.

Therefore, within the added delay time Tt obtained by adding the delay times of the respective delay devices, as many non-periodic reverberant sounds as the number of stages of delay devices are generated. Furthermore, the input signal S, with a larger delay time Td than the addition delay time T,
is delayed and attenuated at a constant rate, and then returned to the human power of the unidirectional signal processing section 1. Therefore, within the next addition delay time Tt, reverberant sounds are generated one after another at different time intervals as described above. Since the above-mentioned human-powered signal feedback operation is repeated infinitely, the generation of reverberant sound can be sustained for a long time with the same number of steps as in the case of a conventional FIR filter. Note that in this case, since the signal is attenuated and re-inputted, there is no risk that reverberant sound will diverge. Moreover, the repetition period (delay time ra) of the reverberant sound is lower than that of the conventional II.
Since it can be set much longer than the R filter (FIG. 5), the periodicity of reverberant sound is significantly relaxed.

Thus, in the present invention, in an audio system or the like,
It becomes possible to generate reverberant sound with little periodicity and a long reverberation time with a limited number of steps.

〔Example〕

FIG. 2 is a circuit diagram showing an embodiment based on the principle of the present invention. Here, one F is used as the unidirectional signal processing unit 1.
An IR filter 11 is provided. This FIR filter 1
1, n delay devices having preset delay times T1 to T, are connected in series. Further, n multipliers 13-1 to 13-n having preset multiplication coefficients g1 to gn are applied to the outputs of the delay devices 12-1 to 12-n.
are connected to each other. Furthermore, the multiplier 13-
An adder 14 is connected to the outputs of 1 to 13-n. In this case, each signal delayed by a predetermined delay time by each of the delay devices described above is amplitude-adjusted by a corresponding multiplier, and then outputted via one adder 14 as an output signal S. The output will be closed. Therefore, this output signal S. When reproduced by a speaker or the like, reverberation sounds are generated sequentially n times corresponding to the number of stages of delay devices.

Further, as the signal delay section 2 (FIG. 1), a delay time Td
A delay device 20 is provided. In this case, in order to prevent the reverberant sounds from overlapping and interfering with each other, the delay time T of the delay unit 20 is set to be larger than the sum of the delay times in the FIR filter 11 (excluding the initial delay time T1). It is preferable to set it to a value. That is, it is preferable to set Td>T2 +T3+...+TI, (1) (in Fig. 2, T, ≧T, +T
2 +T3+...+T9). Furthermore, a multiplier 30 is provided as the signal attenuator 3 (FIG. 1), which multiplies the delayed signal Sd by a constant multiplication coefficient. The multiplication coefficient g of this multiplier 300 is preferably set to a value smaller than 1 in order to prevent the reverberant sound from diverging. That is, g<1 (2)
It is preferable to set it so that The signal that has been subjected to delay processing and multiplication processing (attenuation) so as to satisfy the relationships of equations (1) and (2) above is passed through adder 15 to F
It is fed back to the input of the IR filter 11. That is, the delay device 20, the multiplier 30, and the adder 15 constitute an IIR filter having a specific delay time TaJs and a multiplication coefficient g. Since the input signal S1 is fed back any number of times by this IIR filter, it is possible to generate reverberation sound an infinite number of times, which is determined by the number of steps in the FIR filter 11.

Based on the principle of the present invention (according to the embodiment), the FIR filter 11 is
Since the delay VJT, ~T, and the multiplication coefficients g1~g, can be set, as shown in Figure 3, there is much less periodicity than in the past (Figures 5 and 6).
Moreover, it becomes possible to generate reverberant sound with a long reverberation time in response to a user's request.

FIG. 4 is a diagram showing an example in which the configuration of FIG. 2 is applied to an audio system. Here, the digital filter 10
(see Figure 1) to DSP (DigitalSigna)
This is realized by 1 Processor) 5.

The Kono DSP 5 can process all operations such as multiplication and addition at high speed in real time. Furthermore, the above D
Means for setting coefficients of each element such as a delay device and a multiplier in the SP 5 is realized by a CPU 6 such as a microcomputer. In this case, a digital audio signal obtained by A/D converting an analog audio signal from a source 7 such as a cassette tape recorder or a record player by an A/D converter 8 is used as an input signal Si to the DS.
It is entered in P6. The DSP 6 generates a digital output signal S0 having desired reverberation characteristics by processing the input signal S1 according to a preset delay time, multiplication coefficient, etc. If this digital output signal S0 is converted back into an analog audio signal by the D/A converter 18 and then inputted to the speaker 19 via the amplifier 9, the speaker 19 will finally produce a reverberant sound that suits the user's taste. is reproduced, making it possible to quickly generate the desired sound field.

〔Effect of the invention〕

As explained above, according to the present invention, it is possible to generate reverberant sound with little periodicity and a long reverberation time in an audio system, etc., when the number of steps of the program in the CPU and DSF is limited. A possible signal processing device is realized.

As a result, it becomes possible to quickly set up a sound field having reverberation characteristics that fully satisfy the user's requirements using the reverberant sound.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing the configuration of the first principle of the present invention, Fig. 2 is a circuit diagram showing an embodiment based on the principle of the present invention, and Fig. 3 is a reverberation generated by the signal processing device of the present invention. Figure 4 is a diagram showing an example of applying the configuration of Figure 2 to an audio system; Figure 5 is a diagram showing reverberant sound generated by an IIR filter; Figure 6 is a diagram showing reverberant sound generated by an FIR filter. FIG. In the figure, 1... unidirectional signal processing unit, 2... signal delay unit, 3... signal attenuation unit, 5...
...DSF, 6...CPU. 10...Digital filter.

Claims (1)

[Claims] 1. Processing the input signal (S_i) to obtain a predetermined output signal (S_i)
In a signal processing device having a digital filter for generating a a unidirectional signal processing section (1) to be processed, connected to an input of the unidirectional signal processing section (1), and
The input signal (S_
a signal delay unit (2) having a delay time larger than the sum of the delay times of i); and a delayed signal (S_d) obtained by the signal delay unit (2).
a signal attenuation unit (3) for attenuating the signal by a predetermined ratio and feeding it back to the input of the unidirectional signal processing unit (1);
) A signal processing device comprising: