KR19990064398A - Virtual Sound Field Generation Method for Realistic 3D Sound Implementation - Google Patents

Abstract

The present invention relates to a virtual sound field generation method for realistic stereoscopic implementation.

The present invention includes a first step of inputting a spatial parameter after loading a sound source file; A second step of calculating the reflection intensity of the reflected sound according to the spatial parameter inputted from the first step, and then interpolating between the reflected sounds to improve the sound density; Calculating a reverberation time according to the spatial parameter input from the first step, and then determining a cutoff frequency of the low pass filter suitable for the high frequency absorption time; When the filtering is applied to the sound source from the third step, the sound source is multi-divided into blocks of a certain length in order to avoid a large amount of computation and waste of memory space, and then perform independent operations on each block to maximize efficiency. A fourth step; And a fifth step of harmoniously positioning the reflections implemented from the second and fourth steps to generate a shock response function of the overall space, and is required for the production of CD titles, audio information on the Internet, virtual reality, and game software. Improved sound effects through signal processing algorithms to create a virtual sound field for realistic stereophonic sound realization that enables users to listen to more realistic sounds by providing realism that is difficult for ordinary users to feel in simple mono sound or stereo sound To present.

Description

Virtual Sound Field Generation Method for Realistic 3D Sound Implementation

The present invention improves the sound effects required for the production of CD titles, audio information on the Internet, virtual reality, and game software through signal processing algorithms, thereby providing a realistic feeling that is difficult for ordinary users to feel in simple mono sounds or stereo sounds. The present invention relates to a virtual sound field generation method for realistic stereoscopic sound realization capable of listening to sound.

In general, the sound produced in a particular place varies perception to the listener according to the characteristics of the space, which is determined by the size of the space, the sound absorption rate of the materials constituting the walls, floors, and ceilings, and the variable position of the sound source and the listener. Can be. In order to realize such an effect, conventionally, various filters are combined to model a signal approaching a shock response function of a desired space, or, unlike simulation, in order to consider subtle effects that may occur under real circumstances, It has been used to apply the measured shock response function directly to the sound source using two kinds of acoustic measurement equipment. However, in the latter case, it can be said that it is very difficult to obtain actual measurement data on a place such as a valley in addition to the cumbersome measurement and the number of places to be measured in order to realize an effect corresponding to a desired place. On the other hand, the former has the advantage that it can be realized simply and variously according to user's preference by simulating the presence of various spaces using various filters using only program technique. That is, in this case, it can be said that it is relatively simple compared to the latter case because the infinite effect can be expected depending on the type and number of filters used, the specification of each filter, the combination method between the filters, and the like. Accordingly, the necessity of creating a more natural and realistic sound field effect by adding various algorithms that cannot be seen in the results of the conventional filtering method while implementing the required sound field effect by the combination of the filters has emerged.

Accordingly, an object of the present invention is to create a virtual sound field for realistic stereophonic sound reproduction that can increase the immersion while allowing the listener to create an illusion as if the listener is actually located in an arbitrary space such as a virtual concert hall or a cave. To provide a method.

According to an aspect of the present invention, there is provided a virtual sound field generation method for stereoscopic sound implementation, comprising: a first step of loading a sound file and inputting a spatial parameter; A second step of calculating the reflection intensity of the reflected sound according to the spatial parameter inputted from the first step, and then interpolating between the reflected sounds to improve the sound density; Calculating a reverberation time according to the spatial parameter input from the first step, and then determining a cutoff frequency of the low pass filter suitable for the high frequency absorption time; When the filtering is applied to the sound source from the third step, the sound source is multi-divided into blocks of a certain length in order to avoid a large amount of computation and waste of memory space, and then perform independent operations on each block to maximize efficiency. A fourth step; And a fifth step of generating the shock response function of the entire space by harmoniously positioning the reflection sound implemented from the second and fourth steps.

In addition, the present invention comprises a first step of inputting a spatial parameter after loading a sound source file; A second step of calculating the reflection intensity of the reflected sound according to the spatial parameter inputted from the first step, and then interpolating between the reflected sounds to improve the sound density; Calculating a reverberation time according to the spatial parameter input from the first step, and then determining a cutoff frequency of the low pass filter suitable for the high frequency absorption time; When the filtering is applied to the sound source from the third step, the sound source is multi-divided into blocks of a certain length in order to avoid a large amount of computation and waste of memory space, and then perform independent operations on each block to maximize efficiency. A fourth step; And a computer readable recording medium having recorded thereon a program for executing the fifth step of harmoniously positioning the reflections implemented from the second and fourth steps to generate the overall shock response function.

1 is a flow diagram of a virtual sound field generation procedure according to the present invention.

Figure 2 is a block diagram of a virtual sound field generation block diagram in accordance with the present invention.

Hereinafter, with reference to the accompanying drawings will be described in detail the present invention.

1 is a flow diagram of a virtual sound field generation procedure according to the present invention, Figure 2 is a block diagram of a virtual sound field generation block diagram according to the present invention.

Referring to Figures 1 and 2, the sound field effect processing sequence according to the present invention after loading (S2) the sound source file (S1) to be applied (S2), the size of the virtual space where the listener is located, Parameters S3 related to sound absorption rates of materials constituting walls, ceilings, and floors, locations of listeners and sound sources in the space, and high frequency absorption time are inputted (S4) to start the actual sound field effect processing. The sound field effect is largely divided into the early reflection sound group and the rear bass sound group, which is generally called reverberation. The parameters of the sound absorption rate of the wall, the location of the sound source and the listener are important factors in determining the initial reflection sound and the size of the space and the high frequency absorption time.

First, using the input parameter information, as shown in FIGS. 1 and 2, the intensity of reflection of the sound when the generated sound source is first perceived by the listener and the six sounds reflected from each wall, ceiling, and floor is calculated (S5, 100 and 101). The six early reflections obtained in this way have a dense characteristic in terms of sound density compared with the actual case, so that the density is increased to some extent by interpolation between each reflection sound to overcome this (S6). , 102).

The modeling of the posterior-based sound that lags behind the early reflections is mainly implemented by filtering the sound source. First, a reverberation time S7 corresponding to a space size parameter is calculated and a cutoff frequency of a low pass filter (LPF) suitable for the high frequency absorption time is determined (S8).

The sound source is then transformed to meet the desired purpose by passing through several filters determined according to these two internally calculated factors. The gain 104 of the comb filter S9, 103 having a variable value is fixed by the reverberation time to produce a response curve that approximates the shock response function of the real space, which is attenuated in the form of exponential. , Allpass filter (S11, 105) is to increase the density of the response curve thus made through the gain 106 and the delay to improve the texture of the sound. However, this sound field effect is achieved by complicated signal processing. For this reason, in the case of a general sound source mixed with low frequency or high frequency, many metallic sounds after processing due to the coloration of sound inevitably occurring during the modeling process of gain effect to decay a certain time between the sound sources required for reverberation processing and attenuation. Will be implied. In order to remove the metallic sound of the high frequency component so that the listener can feel the softness more, the Comb filtered sound must be passed through the Lowpass filter (S10, 107). It is necessary to remove the components. Accordingly, as shown in FIG. 2, a lowpass filter is placed in the feedback loop of the comb filter. In general, the length of the sound source used to apply the sound effect is more than several seconds, which is a considerable amount in terms of the number of samples. Therefore, when the aforementioned algorithm is applied to the entire sound source at once, not only does a large amount of calculation be required for the desired signal processing internally, but also requires a large amount of space in terms of memory. Therefore, for effective processing, the long sound source is divided into blocks of a certain length to form a series of block sequences, and then the process is performed by repeatedly processing and summing sound field effects for each block until the last block is reached. (S12).

As described above, after the two reflection sound groups mentioned above are modeled independently for the same sound source, the initial reflection sound and the back-based sound are generated. Is done. That is, since the early reflection sound and the posterior sound are distinguished by temporal differences, among the many reflections that form the early reflection sound group when the sound source is generated, it is reflected by the four walls and finally processed after the sound arrives to the listener, Create an overall shock response curve. This process corresponds to (S13, 108) of FIG.

The present invention described above is capable of various substitutions, modifications, and changes without departing from the spirit of the present invention for those skilled in the art to which the present invention pertains. It is not limited.

As described above, according to the present invention, by effectively implementing the sound field effect, when the listener listens to the sound, it has an excellent effect of causing an illusion as if it is located in a virtual arbitrary space.

Claims (5)

A first step of inputting a spatial parameter after loading a sound source file; A second step of calculating the reflection intensity of the reflected sound according to the spatial parameter inputted from the first step, and then interpolating between the reflected sounds to improve the sound density; Calculating a reverberation time according to the spatial parameter input from the first step, and then determining a cutoff frequency of the low pass filter suitable for the high frequency absorption time; When the filtering is applied to the sound source from the third step, the sound source is multi-divided into blocks of a certain length in order to avoid a large amount of computation and waste of memory space, and then perform independent operations on each block to maximize efficiency. A fourth step; And a fifth step of harmoniously positioning the reflections implemented from the second and fourth steps to generate an impact response function of the entire space. The method of claim 1, The spatial parameter is a parameter related to the size of the virtual space where the listener is located, the sound absorption rate of the material constituting the walls, the ceiling and the floor, the position of the listener and the sound source in the space, and the high frequency absorption time. How to create a virtual sound field for implementation. The method of claim 1, The fourth step is After determining the cutoff frequency of the low pass filter, the long sound source is divided into blocks of a certain length to form a series of block sequences, and the sound field effects are repeated and summed for each block until the last block is reached. Making; When all of the above steps are performed, the method of generating a virtual sound field for stereophonic sound, characterized in that it comprises the step of performing a global pass filter processing to improve the texture of the sound. The method of claim 3, wherein The process of iterating and summing the sound field effect When the listener hears the result sound before performing the step, the comb filter process and the low pass filter process must be performed to remove the high frequency components so as to feel more softness. Way. A first step of inputting a spatial parameter after loading a sound source file; A second step of calculating the reflection intensity of the reflected sound according to the spatial parameter inputted from the first step, and then interpolating between the reflected sounds to improve the sound density; Calculating a reverberation time according to the spatial parameter input from the first step, and then determining a cutoff frequency of the low pass filter suitable for the high frequency absorption time; When the filtering is applied to the sound source from the third step, the sound source is multi-divided into blocks of a certain length in order to avoid a large amount of computation and waste of memory space, and then perform independent operations on each block to maximize efficiency. A fourth step; A computer-readable recording medium having recorded thereon a program for executing the fifth step of harmoniously positioning the reflections implemented from the second and fourth steps to generate an overall shock response function.