JPH05344599A - Automatic discriminating device for adding sound effect - Google Patents

Abstract

PURPOSE:To accurately discriminate a music signal and an announce signal in a frame by discriminating the number by signals of all stereo signals in the frame. CONSTITUTION:On the frame discriminated as a sound frame by a silence number discriminating means 4, a means 15 by level difference numbers discriminates whether the all signals in the frame is a music signal or an announce signal by the level difference between an L signal input 1 and an R signal input 2 of the stereo signal. The one with the higher discriminated signal numbers is decided as a music frame or an announce frame. Thus, the discrimination accuracy is improved and the discrimination whether or not the input is the music or announce signal is accurately performed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic effect sound addition discriminating apparatus for discriminating broadcast contents into music and announcements in order to add reflected sound, reverberation and other sound effects only to received music broadcasts. In particular, it is configured to increase the accuracy of this determination.

[0002]

2. Description of the Related Art In recent years, it has become widespread to provide home audio and car audio with a sound effect addition function, and when the received FM broadcast is music, sound effect is added to announce news. An automatic effect sound addition discriminating apparatus has been devised which automatically discriminates broadcast contents so that no sound effect is added in the case of a program.

The inventor of the present invention has filed an application for an improvement of this automatic effect sound discrimination apparatus as Japanese Patent Application No. 4-17420.

FIG. 5 shows the configuration of the effect sound addition automatic discrimination apparatus according to this prior application. The stereo signal digitized at each sampling period is input to the device by adopting a frame structure including a plurality of them.

This device uses a signal on the left side of the stereo signal.
(L signal) input L signal input terminal 1 and the right signal
R signal input terminal 2 for inputting (R signal), addition processing unit 3 for adding L signal and R signal, “absolute value of L signal + absolute value of R signal” of stereo signal included in frame for each fixed section The number of silent signals contained in the frame is detected, and the number of silent signals is determined by the number of silent signals, and the number of silent signals is determined by the number of silent signals. Any one stereo signal is extracted as a sample from the frame determined to be a voiced frame, and the voiced frame is either a music signal or an announcement signal based on the absolute value of the level difference between the L signal and the R signal of the sample. Level difference discriminating means 5 for discriminating whether or not, silent signal threshold value setting means 6 for determining a silent signal threshold value used for comparison in the silent frequency determining means 4, and a sound effect such as a reflected wave or reverberation. A sound effect generating means 7, a sound effect volume 8 for setting a sound effect level, and output signal addition processing units 9, 10 for adding the sound effect level set by the sound effect volume 8 and the L signal or the R signal, L that outputs a signal
A signal output terminal 11 and an R signal output terminal 12 are provided.

In this automatic effect sound discrimination apparatus,
The L signal and the R signal of the stereo signal digitized at each sampling cycle are added by the input signal addition processing unit 3, and a sound effect such as a reflected sound or a reverberation is generated based on the added input signal. Is generated by. The volume of this sound effect is set by the sound effect volume 8, and then, in the output addition processing units 9 and 10, it is added to each of the L signal and the R signal and output.

The volume of the sound effect volume 8 is set on the basis of the results of the judgments by the silence count judging means 4 and the level difference judging means 5.

The silent signal threshold value setting means 6 provides the silent frequency determination means 4 with a silent signal threshold value to be compared when determining whether the input signal is silent or voiced.
This silent signal threshold value needs to be set in consideration of the input level so that accurate determination can be performed even when the input signal level fluctuates.

Therefore, in the silent signal threshold value setting means 6, the absolute value of the sum of the L signal and the R signal of the sound signal used for detecting the level difference in the level difference determining means 5 is obtained, and the value is 1 / Α (where α> 1) is the silent signal estimation threshold value of the next frame, and the silence signal threshold value of the next frame is set by the following equation from the silence signal estimation threshold value and the silence signal threshold value of the current frame. ing.

K _i = K _i-1 − (K _i-1 −KN) × β (where K: silent signal threshold value, KN: silent signal estimated threshold value,
(β: β <1, i: frame number) Further, the silence number discrimination means 4 and the level difference discrimination means 5 discriminate between the music signal and the announcement signal in the procedure shown in FIGS. 6 and 7,
At the same time, the silence signal threshold value setting means 6 sets the silence signal threshold value of the next frame in the procedure shown in FIG. These procedures will be described below. First, the silence count determination means 4
In FIG. 6, in step 1, the stereo signal “absolute value of L signal + absolute value of R signal” included in the frame for each fixed section and the silent signal range value set by the silent signal range value setting means 8 are set. By comparison, when the “absolute value of L signal + absolute value of R signal” is equal to or greater than the silent signal range value, it is determined that the stereo signal is sound, and when it is smaller than the silent signal range value, it is determined that there is no sound.

Step 2: When a silent stereo signal is determined, the number of silences is counted by incrementing the number of silences one by one, and this operation is continued until the silence detection of the stereo signal in the frame is completed. To do.
When the silence detection for the frame is completed, step 4; it is determined whether or not the frame is a voiced frame based on the counted number of silences. When the number of silent signals in the frame is less than or equal to a certain value (a), the level difference determining means 3 performs the determination. When the number exceeds a,
Step 5; Further, there are two types of correspondence depending on the number,
When the number exceeds a certain value (b, but b> a), it is determined that there is no input and the number of silences is reset (step 7). If the number is less than or equal to b, the input signal is determined to be an announcement signal, step 6; the sound effect volume 8 is subtracted, and then the number of silences is reset (step 7).

Further, in the level difference judging means 5 (FIG. 7), one of the arbitrary stereo signals is extracted as a sample from the frame judged by the silence number judging means 4 as a sound frame,
Step 8: When the absolute value of the difference between the L signal and the R signal of the signal is a certain value (c) or more, the input signal is determined to be a music signal, and Step 9: The sound effect volume 5 is added.

When the absolute value of the difference between the L signal and the R signal is smaller than c in step 8, step 10;
Further, two correspondences are taken according to the absolute value, and when the absolute value is less than a certain value (d, where d <c), it is judged as an announcement and step 11; the sound effect volume is subtracted.

When the absolute value exceeds d, the level difference determination is terminated as an undeterminable signal.

Further, in the silent signal threshold value setting means 6 (FIG. 4), step 12; the L signal and the R signal of the stereo signal used in the level difference judging means 3 are added, and the absolute value is 1
Multiply / α to obtain the estimated value of the silent signal threshold value of the next frame, step 13; using this estimated value of the silent signal threshold value,
The silence signal threshold value (K _i ) in the next frame is calculated by the above formula, and the silence signal threshold value of the silence count judging means 4 is updated. Then, the process jumps to FIG. 6 to reset the number of silences.

As described above, the automatic sound effect discriminating apparatus according to the invention of the prior application obtains the silent signal estimated threshold value from the voiced frame, and uses the silent signal estimated threshold value to adjust the silent signal threshold value to the input level. Is calculated, and the silence signal range value used in the silence number discrimination unit 4 is updated at any time by the calculated value.

The devices before that used a constant value as the silence signal threshold value, but in comparison with this, the device of the invention of the earlier application does not suffer from the difference in the input signal level, and the number of silence times is not affected. It is possible to detect the silent signal of the discrimination means 4, and it is possible to always obtain a constant discrimination result.

[0018]

However, in this effect sound addition automatic discriminating apparatus, the level difference discriminating means 5 discriminates any one sample extracted from the voiced frame so that the entire frame is a music signal frame. Therefore, it is difficult to accurately discriminate the frame because it is determined whether the frame is an announcement signal frame or an announcement signal frame.

The present invention solves such a problem, and an object of the present invention is to provide a sound effect addition automatic discriminating apparatus capable of discriminating correctly between a music signal and an announcement signal.

[0020]

Therefore, according to the present invention, the number of silent signals of a digitized stereo signal included in a frame is detected to determine whether the frame is a silent frame, a sound frame or a non-input frame. In the effect sound addition automatic discrimination device, which comprises a number of times discrimination means and a level difference discrimination means for discriminating between a music signal and an announcement signal by detecting the level difference between the L signal and the R signal of the stereo signal, the level difference discrimination As a means, for each of the stereo signals included in the frame, the distinction between the music signal and the announcement signal is determined, and the frame is a music frame based on the comparison between the number of the determined music signals and the number of the announcement signals. There is provided a level difference number determination means for determining whether it is an announcement frame or an announcement frame.

[0021]

With this level difference number determination means, it is determined whether all stereo signals in the frame are music signals or announcement signals, and the number determined as a music signal is greater than the number determined as an announcement signal. If so, the frame is determined as a music frame, and if the number of music signals determined is less than the number of announcement signals, the frame is determined as an announcement frame.

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS As shown in FIG. 1, the embodiment of the automatic effect sound automatic discriminating apparatus of the present invention is that all the stereo signals included in the frame judged to be the sound frame by the silence number judging means 4 are music signals. A level difference number discriminating means 15 for discriminating whether or not it is an announcement signal is provided. The level difference number determining means 15 includes an L signal input terminal 1 and an R signal input terminal 2
Is directly connected to. In other respects, it is the same as the configuration of the device according to the prior application (FIG. 5).

The operation of this automatic effect sound discrimination apparatus is basically the same as that of the apparatus of the prior application. This device has
A stereo signal digitized for each sampling period is input in a frame configuration, L signals and R signals of the stereo signal are added in the input signal addition processing unit 3, and reflected sound and reflected sound are generated based on the added input signal. A sound effect such as reverberation is generated by the sound effect generating means 7, and the volume of this sound effect is set in the sound effect volume 8 based on the judgments of the silent frequency judging means 4 and the level difference frequency judging means 15. The sound and each of the L signal and the R signal are added by the output addition processing units 9 and 10, and output from the L signal output terminal 11 and the R signal output terminal 12.

In the silence number judging means 4 of this apparatus,
The operation of determining the number of silences is performed by the procedure shown in.

Step 1: Judge whether or not silence detection is completed for all stereo signals in the frame,
Step 2; if not completed, compare the "absolute value of L signal + absolute value of R signal" of the input stereo signal with the silence signal threshold value provided by the silence signal threshold value setting means 6 to obtain "L".
If "absolute value of signal + absolute value of R signal" exceeds the silence signal threshold, it is judged as sound, and the process returns from step 10 (return) to step 1 and "absolute value of L signal + absolute value of R signal". If the "value" is less than or equal to the silence signal threshold, it is judged as silence,
Step 3: Update the silence count, then step 10
Returns to step 1.

When silence detection is completed for all stereo signals included in the frame, step 4;
The silence count for a frame is compared with a fixed value a, and when the silence count exceeds a, it is determined to be a silence frame, step 5; the sound flag is set to 0, step 6; When the number of silences is a or less,
A voiced frame is determined and a voiced flag is set to 1.

Step 7: With respect to the silent frame, the number of silences is compared with a constant value b (> a) to determine whether or not there is no input, and when the number of silences exceeds b,
Reset the number of silences by judging that there is no input (step 9)
Step 8: When the number of silences is equal to or less than the constant value b, the input signal is determined to be an announcement signal, the sound effect volume 8 is subtracted, and Step 9: the number of silences is reset.

In the level difference number discriminating means 15, FIG.
According to the procedure shown in (1), the operation of discriminating the music signal or the announcement signal is performed based on the level difference between the L signal and the R signal.

Step 1; It is judged whether or not the detection of the music signal or the announcement signal for all the stereo signals in the frame is completed. Step 2; If not completed, the L and R signals of the input stereo signal are judged. The absolute value of the signal difference is compared with a constant value c. If the absolute value exceeds c, it is determined to be a music signal, step 3; the number of music signals (MCNT) is updated, and the process returns from step 6 (return) to step 1.

Step 4; Also, in Step 2,
When the absolute value of the difference between the L signal and the R signal is less than or equal to c, it is determined that the signal is not a music signal, and the absolute value is set to a constant value d (<c).
And whether the stereo signal is an announcement signal by comparing with. When the absolute value is less than or equal to d, it is determined as an announcement signal, and step 5; the number of announcement signals (ACN
T) is updated, and the process returns from step 6 to step 1.

When the absolute value exceeds d in step 4, the signal is neither a music nor an announcement and cannot be judged, and the process returns from step 6 to step 1.

When the detection of the music signal or the announcement signal for all the stereo signals included in the frame is completed in this way, step 7; it is judged whether the voiced flag is 1 or not. Step 8: Compare the number of music signals (MCNT) with the number of announcement signals (ACNT), and Step 9: MCNT> AC
If it is NT, it is judged as a music frame and the sound effect volume is added, and step 10; MCNT ≦ ACN.
When it is T, it is determined to be an announcement frame, and the sound effect volume is subtracted.

Step 11; When the voiced flag is 0 in Step 7, or Steps 9 and 10
When the operation of is finished, MCNT and ACNT are reset, and the process goes to the silent signal range value setting means 6.

The procedure for setting the silence signal threshold value in the silence signal threshold value setting means 6 is substantially the same as the procedure in the prior application, and as shown in FIG. 4, step 12; An arbitrary one stereo signal of a frame determined to be a sound frame is extracted as a sample, and its L signal and R signal are extracted.
The estimated value of the silent signal threshold value of the next frame is obtained by adding the signals and the absolute value thereof is 1 / α (where α> 1), and the silence signal threshold value of the next frame is set to K _i = K _i-1 − (K _i-1 −KN) × β (where K: silent signal threshold value, KN: silent signal estimated threshold value,
β: β <1, i: frame number). Then, jump to (1) in FIG.

As described above, in the sound effect addition automatic discrimination apparatus of the embodiment, it is determined whether all the samples of the stereo signal included in the sound frame are music signals or announcement signals, and the number of music signals is determined. If the number judged as a music signal is large by comparing the number judged as an announcement signal,
The frame is determined to be a music frame, and if there are a large number of announcement signals, it is determined as an announcement frame.

The level difference number discriminating means 15 can of course be applied to a conventional effect sound addition automatic discriminating device using a constant silence signal range value.

Further, in the silence number discriminating means 4 and the level difference number discriminating means 15 in the apparatus of the embodiment, as shown in FIGS. 2 and 3, with respect to the processing procedure, the processing of the sample in the frame and the processing after the end of the frame are carried out. And are separated in the step 1. Therefore, it is possible to shorten the processing time as compared with the case where those processings are not separated.

[0038]

As is apparent from the above description of the embodiments, in the effect-sound-added automatic discriminating apparatus of the present invention, in the discrimination based on the level difference between the L signal and the R signal, the discrimination accuracy can be improved, This makes it possible to more accurately determine whether the input signal is music or an announcement.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of an effect sound addition automatic discrimination apparatus of the present invention,

FIG. 2 is a flow chart for explaining the operation of the silent frequency determination means of the same device,

FIG. 3 is a flow chart for explaining the operation of the level difference number determination means of the same device,

FIG. 4 is a flowchart for explaining the operation of the silent signal threshold value setting means of the same device;

FIG. 5 is a block diagram of a conventional effect sound addition automatic discrimination device,

FIG. 6 is a flow chart for explaining the operation of the silence count determination means of the conventional device;

FIG. 7 is a flowchart for explaining the operation of the level difference determination means of the conventional device.

[Explanation of symbols]

 1 L signal input terminal 2 R signal input terminal 3 Input signal addition processing unit 4 Silence number determination means 5 Level difference determination means 6 Silence signal threshold value setting means 7 Sound effect generation means 8 Sound effect volume 9, 10 Output signal addition processing Part 11 L signal output terminal 12 R signal output terminal 15 Level difference frequency determination means

Claims (1)

[Claims] 1. A silence number determination means for detecting the number of silence signals of a digitized stereo signal contained in a frame to determine whether the frame is a silence frame, a sound frame or a non-input frame, and L of the stereo signal. In a sound effect addition automatic discriminating apparatus including a level difference discriminating means for discriminating between a music signal and an announcement signal by detecting a level difference between the signal and the R signal, the stereo included in the frame is used as the level difference discriminating means. A level difference that determines the distinction between the music signal and the announcement signal for each of the signals and determines the distinction between the music frame and the announcement frame for the frame based on the comparison between the number of the determined music signals and the number of the announcement signals. An effect-sound-added automatic discriminating device characterized by being provided with a frequency discriminating means.