JPH11243456A - Digital audio mixing method - Google Patents

Abstract

(57) [Summary] [Problem] A low-level sound is not inaudible after a high-level sound and a low-level sound are mixed. SOLUTION: An average level of each of input digital audio signals A, B, C is detected by a level detection section 13, and these average levels are compared with a reference level. Is determined to be equal to or less than the reference level, a formula is added in which each sample of the voices A and B is weighted by １ ／, and each sample of the voice C is weighted by 加 算.
The selection is made by the calculation formula selection unit 14, and each sample is substituted into the formula, and the mixing calculation is performed and output (16).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice system in which a plurality of persons utter simultaneously to realize a multi-participation type communication function in a wired digital voice communication system such as a wireless system of a cellular phone and a LAN. And a digital audio mixing method for reproducing the same.

[0002]

2. Description of the Related Art At present, a digital audio mixing method in the communication field defines bit lengths in AD conversion and DA conversion for converting an audio signal into a digital value series and vice versa. Therefore, when a plurality of input signals are added, an overflow may occur. To prevent this, the conventional method employs a method of dividing the added sampling value by the number of input signals.

Here, a conventional calculation method will be described using, for example, three voice sampling data shown in FIGS. 5A, 5B, and 5C. In the figure, the vertical axis indicates sample values, and the horizontal axis indicates time. In the conventional mixing method, a sample value of each audio data is added, and a value obtained by dividing the sum by the number of input signals is used as a mixing value. That is, for example, in FIG.
· C '. This mixing value is obtained by the following equation.

(A '+ b' + c ') / 3 = a' / 3 +
b '/ 3 + c' / 3 This indicates that the mixing value equally divides the added sample value regardless of the level of each sound level.

[0005]

In the conventional method, since the average value of all input audio signals is used as the mixing value, in the case of low-level audio, in the above example, the audio of FIG. 5C is divided by the number of signals. In many cases, the voice level is lower than the voice level that can be heard by the user, which makes it difficult to identify the voice.

That is, in the conventional method, there is a problem that it is difficult to hear low-level sound at the stage of input because the level is further lowered after mixing.

[0007]

In order to solve the above-mentioned problem, according to the present invention, the level of each input audio signal is detected, and a predetermined calculation is performed in accordance with the detected audio level. An optimum method is selected from the methods, and an input audio signal is substituted into the selected calculation method to obtain a mixing output.

[0010] With this configuration, a plurality of input audio signals are optimally mixed in accordance with their levels, so that a balanced mixed audio can be generated.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a functional configuration of an audio mixing apparatus to which one embodiment of the present invention is applied. A plurality of, in this example, three audio input units 11A, 11B, and 11C receive audio signals of each channel, and these are stored in a buffer as digital audio signals by the digital audio input processing unit 12, respectively. The level of each digital audio signal input by the level detector 13 is detected. The calculation formula selection unit 14 selects an optimum calculation formula from a plurality of calculation formulas stored in advance in the calculation formula storage unit 15 according to the detection level of the input audio signal. Then, the mixing calculation section 16 performs a calculation process on the input digital audio signal using the selected calculation formula, and outputs the calculated mixing value.

The level detection unit 13 determines the input audio level by comparing one or a plurality of provided reference values with an average of sample values of each input digital audio signal for a certain period of time. Based on the determination result, an appropriate mixing calculation method is selected from a plurality of calculation formulas determined in advance. In this selection method, for example, when the difference between the large and small levels determined by the level detection unit 13 is large, the weight is increased so that the audio data of the small level is not further lost, and the difference is not so large. In this case, a calculation formula that does not make much difference in weighting is selected.

A specific description will be given with reference to FIG. FIG.
The absolute values of the input audio waveform data shown in A, B, and C are respectively shown in FIGS. 2A, 2B, and 2C, and the horizontal broken line in FIGS. 2A, 2B, and 2C indicates the average value, that is, the detected audio level, and the horizontal solid line. Indicates a reference value determined in advance for determining the audio level. Here, the average value is compared with the reference value. If the average value is larger than the reference value, it is regarded as a large waveform, and if it is smaller, it is regarded as a small waveform.

In the case of FIG. 2, FIGS. 2A and 2B show large waveforms, and FIG. 2C shows small waveforms. In this example, a mixing output is obtained for each sample using the following formula. That is, the average of the average values a 'and b' of the voices of FIGS. 2A and 2B belonging to the large waveform is calculated, and the average of this average value and the average value c 'of the voice of FIG. 2C belonging to the small waveform is calculated. That is, the following equation is calculated.

{(A '+ b') / 2} + c '} / 2 =
a '/ 4 + b' / 4 + c '/ 2 Attention is now focused on the lowest level audio data c'. In the conventional method, the sound level of c 'is 1/3 of the level before mixing. However, in this embodiment, at the calculation stage, weighting is performed for each audio level so as to reduce the level difference, so that the decrease in the audio level of c ′ is reduced by 1%.
/ 2, and the level of the large waveform sound is reduced to a '/ 4 and b' / 4, respectively.
For this reason, the audio of the small waveform (FIG. 2C) has higher audio level information than the conventional method in comparison with the waveforms of FIGS. 2A and 2B.

[0014] This is weighted so that the difference between the loud sound and the small sound is suppressed as much as possible so that the loud sound is kept low at the stage of mixing, and the loud sound is kept as low as possible. That's why. The ratio of the mixing value obtained by the conventional method to the entire mixing value of the waveform of FIG. 2C is shown in FIG.
FIG. 3A shows the ratio of the waveform of FIG. 2C in the mixing value obtained in the above embodiment to the entire mixing value, and FIG. In FIGS. 3, A and B, the black portions occupy the entire mixing value, which is the ratio in FIG. It can be seen that the occupation ratio of the waveform in FIG. 2C is greater in FIG. 3B than in FIG. 3A.

As described above, according to this formula, the degree of the lowering of the level of the low sound can be reduced by mixing the sound data with the sound data lower than the reference value and the high sound data while changing the weight. That is, by controlling the level after mixing according to the sound level, all sounds can be easily heard on average. As is apparent from the above description, in the present invention, as shown in FIG. 4, a plurality of digital audio data are input within a predetermined time, processed in parallel for each channel, stored in a buffer, and subjected to input processing ( S1) A level determination process is performed by comparing the average of sample values of each channel for a certain period of time with a reference value (S2), and a calculation formula is selected based on the level determination result (S3). According to the selected formula, a mixing calculation process is performed on the input digital audio data (S4), and the calculation result is output as a signal (S5).

According to the calculation method of this embodiment, before performing the mixing process, each of the didal audio information is divided into large and small levels based on a reference value, and the weight for calculating the mixing value according to the level is changed to achieve a balance. A good sound level can be realized.

[0017]

As described above, according to the present invention, a balanced sound mixing effect independent of the sound level can be obtained as compared with the conventional mixing method. This makes it easy to identify all sounds in the mixed sound.

[Brief description of the drawings]

FIG. 1 is a diagram showing a functional configuration of a mixing device to which the present invention is applied.

FIG. 2 is a diagram showing each sampling data of FIG. 5 as an absolute value.

FIG. 3A is a diagram showing the proportion of FIG. 5C in the whole mixing value obtained by the conventional mixing method, and FIG. 3B is the proportion of FIG. 5C in the whole of the mixing value to which the embodiment of the present invention is applied; FIG.

FIG. 4 is a flowchart showing a processing procedure according to the embodiment of the present invention.

FIG. 5 is a waveform diagram showing various examples of sampling data of an audio signal over a certain period of time.

Claims (1)

[Claims] 1. A digital audio mixing method for superimposing and mixing a plurality of input digital audio signals, wherein each of the plurality of input audio signal levels is detected, and mixing is performed based on the detected audio signal levels. A digital audio mixing method for selecting one of a plurality of calculation methods for calculating a digital signal, and substituting the plurality of input digital audio signals into the selected calculation method to obtain a mixing output.