KR100705930B1 - Apparatus and method for implementing stereophonic - Google Patents

Abstract

The present invention relates to a stereoscopic sound implementation method and apparatus for effectively reproducing stereoscopic sound while reducing the computation process of HRTF (Head Relater Transfer Function) filtering required to implement stereoscopic sound.

According to the present invention, there is provided a method of implementing stereo sound, comprising: modeling attenuation characteristics of direct sound; Modeling a time delay at which the early reflections are reflected and enter the listener's ears; Modeling a time delay between pulse ears at a predetermined symmetry angle at which a spatial feeling of the initial reflection sound is formed; Modeling the attenuation characteristics of the reflected sound; And synthesizing and outputting the direct sound and the reflected sound.

HRTF

Description

Apparatus and method for implementing stereophonic

1 is a view showing a three-dimensional sound implementing apparatus according to an embodiment of the present invention.

2 is a view showing in detail the device for implementing a stereo sound according to an embodiment of the present invention.

3 is a view showing the configuration of a shelving filter according to an embodiment of the present invention.

4 is a flowchart illustrating a method of implementing stereo sound according to an embodiment of the present invention.

The present invention relates to a stereoscopic sound implementation method and apparatus for effectively reproducing stereoscopic sound while reducing the computation process of HRTF (Head Relater Transfer Function) filtering required to implement stereoscopic sound.

Stereo sound is based on the principle that two human ears can hear. From the two signals recognized by both ears, enough information can be extracted to locate the sound source in three-dimensional space. The field of study of these human auditory systems is psychoacoustics.

Stereo sound reproduction aims to digitally reproduce the actual sound field. To do this, it analyzes the sound reaching the listener at different angles and listens to the listening cue, for example, the Interaural Intensity Difference (IID) signal of both ears, and the time of the sound delivered to both ears. It is necessary to recreate the outer ear effect, which acts as a filter on the differential (Interaural Time Difference) signal or the sound reached.

When the sound cues are recreated, they are synthesized to produce stereo sound. At this time, the head transfer function (Head Relater Transfer Function: hereinafter referred to as HRTF).

The HRTF is a filter coefficient modeling a path transmitted from the sound source to the ear drum, and its value varies depending on the relative position between the sound source and the head of the listener.

Normally, HRTF puts microphones in the listener's ears and places the speakers in a location where the location is known, then generates sound through the speakers to record the signals received through the microphones and impulse responses between the original and received signals. The HRTF is obtained by comparing the two-channel system, such as headphones, to be output directly to the listener's ear to reproduce the stereoscopic sound effect.

As described above, a technique of generating three-dimensional stereo sound through HRTF filtering using early reflections is called an OHL technique.

However, conventionally, in order to implement the OHL technique, a problem arises in that a large amount of Imgage source method should be obtained according to the incident angle of the direct sound.

In addition, in order to reduce the above problems, the direct sound direction is changed to implement the Imgage source method offline. Therefore, the time delay, the gain value, and the direction information of the initial reflection sound must be stored in the memory, which requires a large memory capacity. Occurs.

In addition, the HRTF filtering unit of the direct sound for realizing three-dimensional audio must be implemented in real time, which requires five HRTF filtering in each direction of the direct sound, which requires a large amount of computation, and thus requires a real time in a mobile terminal requiring low power. Implementation is impossible.

An object of the present invention is to provide a method and apparatus for implementing stereophonic sound to effectively reproduce stereophonic sound in a low-power portable terminal while reducing the computation process and memory usage of HRTF filtering using early reflection sound to implement stereophonic sound.

According to an aspect of the present invention, there is provided a method of implementing a stereo sound, including: modeling attenuation characteristics of direct sound;

Modeling a time delay at which the early reflections are reflected and enter the listener's ears;

Modeling a time delay between pulse ears at a predetermined symmetry angle at which the spatial reflection of the initial reflection sound is formed;

Modeling the attenuation characteristics of the reflected sound; And synthesizing and outputting the direct sound and the reflected sound.

In addition, the three-dimensional sound implementation apparatus according to an embodiment of the present invention for achieving the above object, HRTF filtering unit for modeling the direction of the direct sound;

A time delay (TD) modeling unit for modeling a time delay of the initial reflection sound;

An ITD modeling unit modeling a time delay according to an incident angle of the initial reflection sound;

It characterized in that it comprises a spectrum generation unit (Spectrum Generation) for modeling the attenuation characteristics of the reflected sound.

Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

1 is a view showing a three-dimensional sound implementation apparatus according to an embodiment of the present invention, Figure 2 is a view showing in detail a three-dimensional sound implementation apparatus according to an embodiment of the present invention.

1 and 2, the HRTF filtering unit 10 for modeling the direction of the direct sound; A time delay (TD) modeling unit 20 for modeling a time delay of the initial reflection sound; An ITD modeling unit 30 modeling a time delay according to the incident angle of the initial reflection sound; Spectrum Generation (40) for modeling the attenuation characteristics of the early reflection sound is configured.

The HRTF filtering unit 10 models the attenuation characteristics of the sound experienced when the direct sound reaches the listener at a specific angle. In other words, it models the impulse response that a sound source experiences in reaching the listener's ear, which allows the sound source to be played at a desired location for 3D audio.

The time delay (TD) modeling unit 20 models a time delay in which the initial reflection sound is reflected and reaches the listener's ear.

In this case, the reflected sound reaches the listener's ear later than the direct sound because the sound source is reflected on the ceiling, floor, and wall and reaches the listener. Therefore, it is modeled as a delay line.

Here, in the present invention, a time delay of 15 to 30 msec is used, and the delay time may be changed according to a system.

The ITD modeling unit 30 generally models the time delay between the left and right ear according to the incident angle of the initial reflection sound because there is no time difference between both ears in O degree, but time difference occurs for all incident angles other than O degree. do.

In the present invention, modeling the time delay between the ear of the incident pulse at the symmetrical angle (90 degrees, -90 degrees) that the spatial desensitization of the initial reflection sound is best, ITD according to the incident angle is calculated by the following equation .

는 귀간의 거리이고 ,

는 소리의 속도를 의미한다.here

Is the distance between ears,

Means the speed of sound.

If a time delay of about 15 samples is generated at 90 degrees, the time delay of 15 samples used at 90 degrees is applied to the ear at -90 degrees, so that no additional calculation is required.

The spectrum generation unit 40 models the attenuation characteristics of the sound that is experienced when the reflected sound reaches the listener at a specific angle.

This can be modeled the direction of the reflection sound through the HRTF filtering of 90 degrees, -90 degrees, in the present invention to approximate the HRTF to the first-order Shelving filter (40-1) as shown in Figure 3 to reduce the amount of computation The system response of the shelving filter is calculated by Equation 2 below.

은 다음의 수학식 3으로 계산되며,

는

로 계산되면, 상기 G는 G in dB for low frequency를 의미한다.here,

Is calculated by the following equation,

Is

When calculated as, G means G in dB for low frequency.

를 가지고 HRTF를 모델링 할 수 있다.In addition, the cutoff frequency parameter for cutting the high frequency is calculated by the following equation (4), and one shelving filter coefficient

You can model HRTF with

Therefore, since the incident angle of the initial reflection sound is fixed at 90 degrees and -90 degrees in any direction of all direct sounds, it is possible to reduce the calculation amount by using only two initial reflection sounds, and to use one filter coefficient using the first shelving filter. Because of the modeling, the operation corresponding to the filtering of the early reflections is replaced by one multiplication operation, which can greatly reduce the amount of computation.

In addition, although the incident delay time, the gain value, and the direction information of the initial reflection sound which are different according to the incidence angle of the direct sound should be stored in the memory, in the present invention, the incidence angle of the initial reflection sound is 90 degrees in any direction of the direct sound. Because it is fixed at -90 degrees, only a minimum of one buffer can be used, which effectively reduces memory usage.

4 is a flowchart illustrating a method of implementing stereoscopic sound according to an exemplary embodiment of the present invention.

Referring to FIG. 4, in order to realize 3D audio, attenuation characteristics of the direct sound reaching the listeners at a specific angle are modeled (S11).

That is, the HRTF filtering unit 10 models the sound attenuation characteristics experienced to reach the listeners at an angle of 90 degrees and -90 degrees in any direction of the direct sound.

Afterwards, the time delay between the reflection of the initial reflection sound and the listener's ear is modeled (S12), and the time delay between the ear delay of the symmetrical angle (90 degrees, -90 degrees), which is the best spatial shaping with the initial reflection sound. (S13).

That is, the time delay (TD) modeling unit 20 models the time delay in which the initial reflection sound is reflected and reaches the listener's ear, and in the ITD modeling unit 30, the symmetry angle at which spatial feeling is best formed as the initial reflection sound is (90). Model the time delay between the pulses of the pulse (Fig.

Therefore, unlike the conventional method of applying the four initial reflections, the present invention reduces the computational amount compared to the conventional method because the modeling of the reflections of the symmetry angle (90 degrees, -90 degrees), which is a good spatial formation of the initial reflection sounds.

In addition, since the incident angle of the initial reflection sound is fixed at 90 degrees or -90 degrees in any direction of all direct sounds, only a single buffer can be used, thereby effectively reducing memory usage.

Thereafter, the attenuation characteristic of the sound that is experienced when the reflected sound reaches the listener at a specific angle is modeled (S14).

The attenuation characteristic of the reflected sound is made through the spectrum generator 40, and the direct sound modeled by the HRTF filtering unit 10 and the reflected sound output from the spectrum generator 40 are left and right. The signal R is separately synthesized and finally output as signals of two channels (S15).

As described above, the present invention has the advantage of outputting three-dimensional audio while reducing the amount of computation and memory usage of the HRTF filtering portion of the OHL technique using the early reflection sound.

Since the OHL technique is already known or used by those skilled in the art to which the present invention pertains, a detailed description thereof will be omitted.

So far, the present invention has been described with reference to the embodiments, and those skilled in the art to which the present invention pertains may implement embodiments of the present invention in a different form from the detailed description of the present invention within the essential technical scope of the present invention. Could be. Here, the essential technical scope of the present invention is shown in the claims, and all differences within the equivalent range will be construed as being included in the present invention.

According to the method and apparatus for implementing the stereo sound according to the present invention, the amount of computation is reduced by using the first shelving filter, and the unnecessary angle of memory is fixed by fixing the incident angle of the initial reflection sound by 90 degrees and -90 degrees in any direction of all direct sounds. It is possible to effectively implement three-dimensional stereophonic sound in a low-power portable terminal while reducing the cost.

Claims (7)

Modeling attenuation characteristics of the direct sound; Modeling a time delay at which the early reflections are reflected and enter the listener's ears; Modeling a time delay between pulse ears at a predetermined symmetry angle at which a spatial feeling of the initial reflection sound is formed; Modeling the attenuation characteristics of the reflected sound; And synthesizing the direct sound and the reflected sound and outputting the synthesized sound. The method of claim 1, The time delay modeling of the pulse ear modeling the direction of the reflected sound through the HRTF filtering of 90 degrees, -90 degrees symmetry angle, the symmetry angle is characterized in that the changeable. The method of claim 1, 3. The method of claim 3, wherein the initial reflection sound incident at the predetermined symmetry angle is delayed by using a predetermined time delay value. HRTF filtering unit for modeling the direction of the direct sound; A time delay (TD) modeling unit for modeling a time delay of the initial reflection sound; An ITD modeling unit modeling a time delay according to an incident angle of the initial reflection sound; And a spectrum generation unit for modeling the attenuation characteristics of the reflected sound. The method of claim 4, wherein The ITD modeling unit delays the time delay of the initial reflection sound according to a predetermined incident angle.

, (

Is the distance between ears,

Is a speed of sound). The method of claim 4, wherein The spectrum generation unit (3) is a stereoscopic sound generation apparatus, characterized in that for modeling the reflected sound using a shelving filter (Shelving filter). The method of claim 6, The shelf filter coefficient is

Apparatus for stereophonic sound, characterized in that calculated by the formula.

Images