Reverberation sound generating apparatus

Abstract

 A reverberation sound generating apparatus, comprising a plurality of comb filters (1₁-1_n) each having a delay element and a coefficient multiplier, and an adder (2) for adding output signals of the comb filters, wherein the numbers of the samples corresponding to each delay time of the comb filters are mutually prime. With such arrangement, it is possible that the impulse responses of each of the comb filters do not coincide with each other almost at all within the reverberation time. Thus, reverberation sound is not deviated over time and is dispersed uniformly, providing satisfactory reverberation tone quality.

Description

[0001] The present invention relates to a reverberation sound generating apparatus using a plurality of comb filters.

[0002] A reverberation sound generating apparatus is known, which adds reverberation effect to audio signal sound. As the conventional reverberation sound generating apparatus, there is known an apparatus for generating reverberation sound by connecting in parallel a plurality of comb filters which have different delay time characteristics and by mixing output signals of the comb filters. However, such conventional type apparatus is disadvantageous in that the tone quality of the reverberation sound is deteriorated when impulse responses are overlapped between the comb filters.

[0003] It is an object of the present invention to provide a reverberation sound generating apparatus, which can prevent the overlapping of impulse responses between the comb filters and improve the tone quality of the reverberation sound.

[0004] The reverberation sound generating apparatus according to the present invention comprises a plurality of comb filters each having a delay element for delaying input digital audio signals each representing a sample of an audio signal and multiplying means for multiplying a delayed signal generated from the delay element by a coefficient and for feeding back to the input of the delay element, and adding means for adding output signals of the comb filters, wherein the numbers of samples corresponding to each delay time of the comb filters are mutually prime.

[0005] In the reverberation sound generating apparatus according to the present invention, it is possible that the impulse responses of the comb filters do not coincide with each other within the reverberation time because the numbers of the samples corresponding to each delay time of the comb filters are mutually prime. Accordingly, reverberation sound is uniformly dispersed without deviating over time, and good reverberation tone quality can be obtained.

[0006] An embodiment of the invention will now be described by way of example only and with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

[0007] 

Fig. 1 is a block diagram showing an embodiment of this invention;

Fig. 2 is a circuit diagram showing the arrangement of a comb filter; and

Fig. 3 shows a reverberation response to an impulse of a comb filter.

[0008] In a reverberation sound generating apparatus according to the present invention as illustrated in Fig. 1, a plurality of comb filters 1₁ to 1_n ("n" indicates the number of the comb filters) are connected to an input terminal IN. To the output terminal of the comb filters 1₁ to 1_n, an adder 2 is connected. To the output terminal of the adder 2, a plurality of APFs (all pass filters) 3₁ to 3_m ("m" represents the number of the APFs ) are connected serially. The output terminal of the APF 3_m is connected to the output terminal OUT of this apparatus. The APFs 3₁ to 3_m comprise secondary IIR (infinite impulse response) type filters and have group delay characteristics mutually different with constant gain to frequency. The APF is disclosed, for instance, by the Japanese Patent Laid-open Publication No. 62-161210, and detailed description will not be given here.

[0009] As shown in Fig. 2, the comb filters 1₁ to 1_n consist of an adder 6, a delay element 7 and a multiplier 8. Namely, an input signal is sent to the delay element 7 through the adder 6, and it is delayed by the delay time T of the delay element 7. The delayed signal is output and is sent to the multiplier 8 at the same time. The multiplier 8 multiplies the delayed signal by a feedback coefficient g (0 < g < 1), and the output signal of the multiplier 8 is sent to the adder 6 and is added to the input signal. Therefore, when a pulse signal representing amplitude a₀ is supplied as an input signal, the pulse signal a₀ is output as a₁ after the delay time T as given in Fig. 3. Then, it is multiplied by the coefficient g for attenuating, and the attenuated pulse signal a₂ is output after the delay time T has elapsed. By repeating this action, the pulse signals a₃, a₄, a₅, ....... are output after succesive intervals T after the output of the pulse signal a₂, and the signal level is gradually decreased.

[0010] The digital audio signal supplied to the input terminal IN is sent to each of the comb filters 1₁ to 1_n. Accordingly, for each sample value of the input digital audio signals, it is delayed by delay time T₁ to T_n being the respective delay times of filters 1₁ to 1_n. Thereafter, the signal level is gradually decreased as each delay time elapses.

[0011] Each of the output signals of the comb filters 1₁ to 1_n is delayed by the APF 3₁ to 3_m, after being mixed by the adder 2. Because the APF 3₁ to 3_m have different group delay characteristics, signal density is increased by further delaying the output signal of the adder 2. For example, in case of impulse input, sound level is decreased as time elapses, but the sound continues.

[0012] The delay time T₁ to T_n of the delay element of the comb filters 1₁ to 1_n differs from each other and is given as a multiple of the sampling period. The number of the samples corresponding to each delay time, i.e. the number of delay samples, is mutually prime. For example, if a sampling frequency is 44.1 [kHz], the delay time T₁ of the comb filter 1₁ is 1187 in the number of delay samples and 26.9 [ms] in time. The delay time T₂ of the comb filter 1₂ is 1367 in the number of the delay samples and 31.0 [ms] in time. The delay time T₃ of the comb filter 1₃ is 1549 in the number of the delay samples and 35.1 [ms] in time. Because the numbers of the delay samples are mutually prime, the time up to the overlapping of the impulse responses of the comb filter 1₁ and 1_n is very long. By showing this with numerical values, the number of samples between the comb filter 1₁ and 1₂ is: 1187 x 1367 = 1622629, and it is 36.8 [s] in time. This value means that the impulse responses do not overlap each other if the fact that reverberation time is within 2 - 4 seconds is taken into account.

[0013] In the above embodiment, description has been given of the apparatus, which is designed in terms of hardware. The invention is naturally not limited to this, and the comb filters and the APFsmay be arranged by computation through a DSP (digital signal processor).

[0014] As described above, in the reverberation sound generating apparatus according to this invention, the numbers of the samples corresponding to each delay time of a plurality of comb filters are mutually prime, and it is possible that the impulse responses of each comb filter do not coincide with each other almost at all within the reverberation time. Accordingly, the reverberation sound is not deviated over time and is dispersed uniformly, thus giving satisfactory tone quality.

Claims

1. A reverberation sound generating apparatus comprising:
   a plurality of comb filters each having a delay element for delaying input digital audio signals each representing a sample of an audio signal and multiplying means for multiplying a delayed signal generated from said delay element by a coefficient and for feeding back to the input of said delay element, and
   adding means for adding output signals of said comb filters,
   wherein the numbers of the samples corresponding to each delay time of said comb filters are mutually prime.

2. A reverberation sound generating apparatus according to claim 1, wherein all pass filters connected serially are connected to the output of said adding means.

3. A reverberation sound generating apparatus according to claim 1 or 2, wherein the delay times of said delay elements are given as multiples of the sampling period.

Drawing