KR0134625Y1 - Subsequent reverberation sound generating circuit - Google Patents

Abstract

The present invention relates to a rear reverberation sound generating circuit that performs reverberation sound effect processing on the reproduced sound so as to secure a sense of presence in an audio device. It is difficult to secure the realism of sound because it cannot follow the change of period, amplitude and frequency characteristics due to sound, diffuse reflection, and absorption rate difference between each wall surface, and it has only a slight effect to realize three-dimensional and spatial feeling. It is for solving the problem that the reproduction fidelity is lowered.

The present invention secures irregularity in reverberation sound processing by randomly varying the delay time and feedback coefficient according to time in performing reverberation sound processing by a comb filter. It can be provided with high accuracy corresponding to the actual playing scene, and it can be applied to an audio reproducing device to secure high fidelity of playback sound.

Description

Rear Reverberation Generator Circuit of Audio Equipment

1 is listening sensitivity response characteristics in the sound source scene.

2 is a conventional rear reverberation sound generating circuit diagram.

(A), (b), (c) of FIG. 3 are rear reverberation frequency characteristic diagrams by a conventional circuit.

Figure 4 (a) is a random coefficient output circuit in the rear reverberation sound generating circuit of the present invention.

(b) is a variable comb filter circuit in the rear reverberation sound generating circuit of the present invention.

5 is a random coefficient generator circuit in the random coefficient output circuit of the present invention.

(A), (b) and (c) of FIG. 6 are rear reverberation frequency characteristic diagrams of the present invention.

* Explanation of symbols for main parts of the drawings

5: Random coefficient generator 6: Feedback gain memory

7-1 to 7-k: Gain adder 8: Delay time memory

9-1 to 9-k: delay adder 10-1 to 10-k: delay adder

11-1 to 11-k: Multiplier 12-1 to 12-k: Input adder

13: output adder

The present invention relates to a rear reverberation sound generating circuit which performs reverberation sound processing on the reproduced sound so as to secure the presence of sound in an audio device.

In general, the realism (a stereoscopic sense of sound and a sense of space) that a listener feels in a real performance scene is difficult to exclude the influence of the rear reverberation and the reflection of reflection on the wall, besides the sound directly transmitted to the listener from a sound source (such as a musical instrument or a singer).

Figure 1 shows the impulse response characteristic of the rear reverberation sound, whereby the initial reflection sound (50 to 80 msec) is good enough to reproduce the frequency characteristic, but the subsequent reverberation sound (Subsequent Reverberation Sound) is frequency characteristic. This configuration is very irregular and difficult to reproduce. However, since the rear reverberation sound acts as an important factor in feeling the presence of sound, the rear reverberation effect processing is conventionally performed by using a circuit in which a plurality of comb filters are connected in parallel. In one circuit configuration, referring to FIG. 2, delays 1-1 to 1-k delay the input signal x (n) by a predetermined time Nk, and feedback gains g1 to gk to the delay signal. Multiply the multipliers (2-1 to 2-k) by the multiplier, the input adders (3-1 to 3-k) to add the multiplied signal and the original input signal, and the added output signal to synthesize the final output signal ( and an output adder 4 outputting to y (n)). Therefore, in the conventional rear reverberation sound generating circuit, the input signals x (n) are set to the predetermined times N1, N2, N3, in the respective delayers 1-1 to 1-k having different delay times. After delaying by .NK, the feedback gains gk are multiplied by multipliers 2-1 to 2-k having different feedback gains g1, g2, g3, ..., gk. The multiplied feedback signal is again added to the input signal x (n) by the input adders 3-1 to 3-k, and then the signals C1 (z) and C2 (z) which are added in the comb filter. , C3 (z), ..., CK (z) are added by the output adder 4, and finally the audio signal y (n) subjected to the rear reverberation sound effect is output.

In this case, the result of observing the frequency characteristic using the case of x (n) = impulse is shown in (a), (b), (c) of FIG.

That is, k = 1, the first comb filter (1-1) (2-1) (3-1) is between 1 / (1-g1) and 1 / (1 + g1) as shown in FIG. It repeats regularly and its period is 1 / N1t, depending on the delay time N1.

Therefore, in the k-th comb filter (1-k) (2-k) (3-k), 1 / Nkt between 1 / (1-gk) and 1 / (1 + gk) as shown in (b) of FIG. The audio signal is repeated in cycles, and the result of synthesizing such signals in the output adder 4 is the rear reverberated audio signal (Y (z); z-Transform) as shown in FIG. Is output as.

However, according to the conventional rear reverberation sound generating circuit, it is impossible to realize this by following the change of the period, amplitude, and frequency characteristics caused by the irregular reflection and the irregular reflection, the absorption difference between each wall surface. Therefore, it is difficult to secure a sense of presence of the sound, and there is a problem in that the reproduction fidelity is lowered because there is only a slight effect in realizing a three-dimensional feeling and a spatial feeling.

This is because, as in FIGS. 2 and 3, each comb filter maintains a constant value with its delay time (NK) and feedback gain (gk) not variable, and as shown in the frequency characteristic thereof. Although the reverberated audio channel has a constant amplitude (VH) (VL) and regularity that is repeated at a periodic period, it is difficult to secure the reverberation effect in a real performance site where irregularities are assumed.

Therefore, the present invention secures irregularities in the reverberation sound processing by randomly varying the delay time and the feedback coefficient according to time in performing the reverberation sound processing by the comb filter. It is possible to provide a high accuracy corresponding to the actual performance scene, and to provide a rear reverberation sound generating circuit of the audio device to ensure a high fidelity of the playback sound, the following (a) and (b) of FIG. Referring to the circuit configuration of the present invention with reference to as follows.

That is, the rear reverberation sound generation circuit of the present invention includes a random coefficient generator 5 in which the feedback gain and the delay time of the comb filter generate a random number (RNG) (RND), and the comb filter feedback gains (g1 to gk). A feedback gain memory 6 in which reference values of the feedback values are stored and sequentially output, a feedback gain memory 6 in which the reference values of the feedback gains g1 to gk are stored and sequentially output, and the feedback gain and a random number. The gain adders 7-1 to 7-k for outputting the variable gains g1 'to gk' and the reference values of the comb filter delay times N1 to NK are stored and sequentially outputted. ), Delay adders 9-1 to 9-K for outputting variable delay times N1 'to NK' by adding the delay time and the random number, and the variable delay times N1 'to NK'. Delayers 10-1 to 10-k that receive and delay the input audio signal x (n) with a variable delay time; Multipliers 11-1 to 11-k that receive variable gains g1 'to gk' and multiply the delayed audio signal with variable gains, and input the signal x (n) to the multiplied audio signal. ), An input adder 12-1 to 12-k for adding processing, and an output adder 13 for adding up the processed signal and outputting the reverberated final audio signal y (n). The rear reverberation sound processing operation will be described with reference to FIGS. 4 to 6 as follows.

First, as shown in FIGS. 4A and 5, the random coefficient generator 5 is supplied with a clock CLK to adjust the variable gains g1 'to gk' and the variable delay times N1 'to NK'. Each random number (RNG) (RND) for output is generated.

That is, the input clock CLK is supplied to the clock of the shift register 5B through the exclusive oragate 5A, and accordingly, the shift register 5B sequentially feeds the feedback data of the input stage I1 to the next stage output stage. Shifting to the input stage 01 / I2, 02B / I3,03 / I4, the data of the final output stage 04B is supplied to the input stage DA of the dual shift register 5C of the rear stage.

The dual shift register 5C shifts the data of the input terminal DA according to the clock supplied to the clock input terminal CLKA (CLKB) and outputs the shifted data to the output terminals Q1A, Q2A, Q3A and Q4A. Then, the data of the output terminal Q4A is supplied to the next input terminal DB again, and the final data is output from the output terminals Q1B, Q2B, Q3B and Q4B as a result of the shift of this data.

That is, the data at the output terminals Q1A, Q2A, Q3A, and Q4A becomes a random number RNG for generating variable gain, and the data at the output terminals Q1B, Q2B, Q3B, and Q4B is variable delayed. It is a random number (RND) for generating time.

The data of the output terminals Q1A and Q4B is fed to the input terminal I1 of the shift register 5B through the exclusive orifices 5D, 5E, and 5F, respectively. As a result, the random number is generated every time the clock CLK is input.

The random numbers RNG (RND) thus output are applied to the gain adders 7-1 to 7-k and the delay adders 9-1 to 9-k, respectively.

In this case, the gain adders 7-1 to 7-k and the delay adders 9-1 to 9-k each include a feedback gain provided with data data in the feedback gain memory 6 and the delay time memory 8, respectively. (g1 to gk) and delay times (N1 to NK) are supplied in turn, so that each adder 7-1 to 7-k (9-1 to 9-k) is assigned a random number (RNG) (RND) and a feed. The variable gains g1 'to gk' and the variable delay times N1 'to NK' that are randomly changed are added by adding the back gains g1 to gk and the delay times N1 to NK.

The variable gains g1 'to gk' and the variable delay times N1 'to NK' are a multiplier 11-1 of delay time and feedback gain by delays 10-1 to 10-k of the comb filter. To 11-k) to randomly change the multiplication gain.

Therefore, reverberation sound processing that is irregular in time and encompasses the feeling of listening at the actual listening site is performed.

That is, as shown in FIG. 6A, k = 1 and the reverberation sound processing results in the first comb filters 10-1, 11-1 and 12-1 have a delay time N1 'and a gain g1. ') Is randomly changed, and the reverberation sound processing is performed in which the amplitude and period are irregular and change, which is also performed by the k-th comb filter (10-k) (11-k) (12-k). Since it is performed as shown in (b) of FIG. 6, the final output signal y (n) summed by the output adder 13 listens to the actual performance sound (or the voice of the singer) as shown in (c) of FIG. 6. Proximity with minimum deviation.

Therefore, according to the present invention, since the rear reverberation sound processing effect can be secured by the listening feeling in the actual performance, the quality of the device as well as the qualitative improvement of the listening feeling in the music can be obtained. Audio playback is possible.

Claims (2)

A feedback gain memory which stores a random coefficient generator 5 for generating a feedback gain of the comb filter and a random number (RNG) (RND) of delay time and a reference value of the comb filter feedback gain (g1 to gk) and sequentially outputs them. (6), the gain adders 7-1 to 7-k for adding the feedback gain and the random number to output the variable gains g1 'to gk', and the comb filter delay times N1 to NK. A delay time memory 8 in which reference values are stored and sequentially output, delay adders 9-1 to 9-k for adding variable delay times and random numbers to output variable delay times N1 'to NK', A delay unit 10-1 to 10-k receiving the variable delay time N1 'to NK' and delaying the audio signal x (n) with a variable delay time, and the variable gains g1 'to multipliers (11-1 to 11-k) provided with a variable gain and multiplied by the delayed audio signal, and the multiplied audio; An output that adds the signal to the input signal x (n), the input adders 12-1 to 12-k for adding the processed signal, and adds the processed signal to output the reverberated final audio signal y (n) A rear reverberation sound generating circuit of an audio device composed of an adder (13). The random coefficient generator (5) according to claim 1, wherein the exclusive coefficient generator (5A) for supplying the input clock (CLK) and a shift register (5B) for shift output of the feedback data in accordance with a clock. And a dual shift register (5C) that outputs shift output data as random number (RNG) (RND) data, and an exclusive ogate (5D) (5E) (5F) that feeds back shift data to generate a random number. Rear reverberation sound generation circuit of audio equipment consisting of