JP3347242B2 - Signal delay time detecting device and audio delay countermeasure monitoring device using the detecting device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to, for example, in a relay broadcast using a communication satellite, a signal delay time generated between an audio signal at a relay site and a signal received from a broadcast station at a relay site. 1. Field of the Invention The present invention relates to a signal delay time detecting device for detecting, and an audio delay countermeasure monitor device using the detecting device.

[0002]

2. Description of the Related Art For example, in a case where a broadcasting station studio and a relay site are interconnected for relay broadcasting, there is a problem due to a delay of an audio signal.
That is, if the announcer at the relay site speaks in front of the microphone while monitoring the audio signal being broadcast from the receiver with the earphone, the user's voice will be heard from the earphone with a delay of the delay time generated on the broadcast line, It can be uncomfortable or, in some cases, interfere with the announcement process.

Conventionally, various countermeasures have been studied to prevent the above adverse effects. For example, Japanese Patent Application Laid-Open No. 5-227
SNG (Satellite News Gat) disclosed in Japanese Patent Publication No. 110
hering) sidetone suppression device delays the output signal of the microphone at the relay site by a fixed delay circuit for a fixed time,
A signal obtained by subtracting the delay audio signal from the on-air audio signal is output to the earphone as a monitor signal on the spot. As a result, the announcer at the site does not hear the voice spoken by himself / herself from the earphone, so that the above-mentioned adverse effect is eliminated.

Since the above-described suppressing device sets a fixed fixed value as the delay time, there is a difference between the delay time and the actual accurate delay time, and the compensation is performed by an adaptive equalizer. However, there is a problem in that stable performance is always obtained.

On the other hand, recently, digital transmission has increased, and in this case, delay times associated with digital conversion and digital compression processing are inevitable. Its value will vary significantly depending on the configuration. Therefore, if the delay processing time is set to a constant value as introduced in the above-mentioned literature, it becomes more difficult to cope with the delay processing time.

Therefore, it is required to measure a deviation between two signals. As an example, for example, "Broadcasting Technology" 1994. 92-95, "Video / Audio Time Shift Measurement Apparatus" is introduced. However, when the video signal and the audio signal, which are not displaced from each other at the program transmission source, are transmitted through the transmission path and received at the receiving end, how much time difference exists between the two signals? This requires complicated processing such as superimposition of an audio signal on a video signal, and does not meet the purpose of measuring the delay time of one audio signal, which is a problem described above.

SUMMARY OF THE INVENTION The present invention meets the above-mentioned problems and provides a signal delay time detection capable of simply and accurately detecting the delay time from only data of an original signal and a signal delayed by a certain time. It is an object of the present invention to obtain a device and a monitor device for countermeasures against audio delay using the detection device.

[0008]

According to a first aspect of the present invention, there is provided a signal delay time detecting apparatus, wherein a signal X relating to a voice detected at a voice generation site and a signal including the signal X are transmitted through a relay means or the like. A signal Y obtained by monitoring a signal transmitted from a broadcasting station is input, and the sound of the signal X included in the signal Y is input.
In a signal delay time detecting device for detecting a delay time for removing a voice component , sampling of data over a predetermined reference time Tr from a predetermined time (hereinafter, referred to as time t = 0) is started for the signal X. Means for storing as fetched data Xi (i = 1 to (Tr / Ts)) at a period Ts, and for the signal Y, starting from the time t = 0 as a starting point, a predetermined delay time predicted maximum value Tm and the reference time Tr
Data over the time (Tm + Tr) obtained by adding the data at the sampling period Ts, and the fetched data Yj (j = 1 to (T
m + Tr) / Ts), and when the time width corresponding to the detection accuracy is Td (Td <Tr), the starting point is sequentially shifted to time t = 0, Td, 2Td,. From the fetched data Yj, and reference data Yi, k ｛i = 1 to (Tr / Ts), k = 1
（(Tm / Td)}, the fetched data Xi is referred to as reference data Xi, and both reference data Xi and Yi,
k, a correlation value Zk (k = 1 to (Tm / T
d)), a means for searching for the largest one among the correlation values Zk, Zp = Zk (k = p), and a value p of k at which the correlation value Zk is maximum is expressed by the following equation. Means for calculating the delay time ΔT, and the reference time Tr
Is set to 0.5 sec or more, and the calculated delay time ΔT is made effective when the searched maximum value Zp of the correlation value is 0.5 or more.

(Equation 2)

According to a second aspect of the present invention, there is provided a signal delay time detecting device.
2. A method according to claim 1, wherein all of the fetched data Xi fetched from the signal X or the fetched data Yj fetched from the signal Y are at or below a predetermined noise level. The data is rewritten by taking in from X and signal Y, and the subsequent operation is executed again.

According to a third aspect of the present invention, there is provided a signal delay time detecting device.
In claim 1 or 2 , when the searched maximum value Zp of the correlation value is less than 0.5 , the signal X and the signal Y are taken in again, the data is rewritten, and the subsequent operation is executed again. It was made.

[0011] sound delay measures monitoring device according to claim 4, the signal delay time detection apparatus according to any one of 請 Motomeko 1 to claim 3, delays the signal X is detected by the signal delay time detection device means for producing a signal XD delayed by time, Oyo by subtracting the signal XD from bicine No. Y, monitor signal to remove audio components of the signal X YM
Is provided.

According to a fifth aspect of the present invention, there is provided an audio delay countermeasure monitoring apparatus according to the fourth aspect, wherein the signal delay time detecting device detects a delay time at a predetermined time interval and uses the detected value to generate a signal XD. The time is sequentially corrected.

[0013]

FIG. 1 is a block diagram showing a schematic configuration of a signal delay time detecting device according to an embodiment of the present invention. For example, assuming a dual-broadcast news program connecting a studio of a broadcasting station and an accident site, a signal X shown in the figure is an audio signal obtained by detecting a sound of a reporter reporting a local situation at the accident site with a microphone. Is applicable.

The audio signal X from the relay site is usually
The contents of the broadcast antenna are transmitted from the transmitter at the relay destination to the broadcasting station, which is the main station, via the communication satellite, and monitored by the studio, and the sound of the announcer in the studio is added in a form that interacts with the local reporter. It is more aired. Then, the on-air radio wave is received by a receiver at a relay site, and the audio signal is supplied as a monitor signal to an earphone of a local reporter. The signal Y in FIG.
Corresponds to this monitor signal.

Further, 1 and 2 are A / D converters for converting the signal X and the signal Y into digital signals, respectively.
Is a central processing unit that performs input / output of each signal and various arithmetic processing,
Reference numeral 4 denotes an input unit such as a keyboard for inputting various setting values and the like. Reference numeral 5 denotes a storage unit including a ROM for storing a software program for performing processing to be described later and a RAM for performing necessary storage in an arithmetic processing process. An output unit such as a display or a printer that outputs a delay time or the like obtained by calculation.

Next, the operation of the signal delay time detecting device will be described in detail. FIG. 2 is a diagram conceptually showing the operation principle, and FIGS. 3 and 4 are flowcharts showing the flow of processing. Now, as shown in FIG. 2, the signal X and this signal Δ
It is assumed that a signal Y delayed by T is input. First, the maximum predicted delay time Tm is input from the input unit 4 (step S1). For this, an appropriate value is set based on past performance data, calculation data, and the like so that the actual delay time does not exceed this Tm. If the set value Tm is too large, the detection accuracy does not decrease, but the time required for detection increases.

Next, the signal X is fetched (step S2).
This capture is performed at the sampling time Ts and the start time t.
Starting from = 0, it is performed for the reference time Tr. The longer the sampling period Ts, the lower the accuracy, and the shorter the sampling period Ts, the longer the operation time. 4K
Sufficient accuracy was obtained by setting the period to about 1/4000 sec corresponding to Hz. The longer the reference time Tr is,
The possibility of erroneous detection is reduced, but the detection response is delayed. From the results of trials conducted by the inventors, it has been found that by setting Tr ≧ 0.5 sec, the chance of erroneous detection can be suppressed to 5% or less.

The fetched data Xi is, after all, its i
Is the sequence data in which is 1 to (Tr / Ts), and Tr / Ts≡n
In other words, the above specific example Tr = 0.5 sec, Ts = 1
By substituting / 4000 sec, n = 2000.

Next, the signal Y is fetched (step S3). This is from t = 0 (Tm), which is the same as the signal X capture start time.
+ Tr) only. The reason why the end time is set to (Tm + Tr) is to enable the detection even when the actual delay time ΔT is equal to Tm.

Next, the acquired data Xi, Yj obtained above
Is determined whether or not can be used as data for the subsequent calculation analysis (step S4). That is, if all of the fetched data Xi is at or below the noise level, or if all of the fetched data Yj is at or below the noise level (step S
4 is YES), which corresponds to the case where no voice is accidentally input within the capture time, and the solution of the delay time cannot be obtained.
Returning to step S2, the capturing process is executed again.

If NO in step S4, the fetched data Xi is regarded as the reference data Xi as it is, and the fetched data Yj
Then, the reference data Yi, k is extracted from (step S6). As shown in FIG. 2, this extraction is performed sequentially from the time t = 0 and shifted by the time width Td over the reference time Tr. In order to sequentially calculate a correlation value Zk to be described later, k = 1 is initially set in step S5, and then the process proceeds to step S6. Therefore, k is 1 to (Tm / T
The value is in the range of d).

Subsequently, the reference data Xi and the extracted 1
A correlation value Zk is calculated from the set of reference data Yi, k by the following equation (step S7).

[0023]

(Equation 3)

This equation can be expressed as both reference data Xi,
Assuming that Yi, k is a linear and stationary signal, a convolution operation (for example, published in CQ Publishing Co., Ltd., August 1989, separate volume interface "Using DSP", pp. 15-20)
(See, for example, Asakura Shoten Co., Ltd., November 1994, “Statistical Library Spectral Analysis”, pp. 25-33). As can be seen from the calculation formula, if the two referenced data Xi and Yi, k have exactly the same characteristics, the correlation value Zk
Becomes 1 and the correlation value Zk decreases as the identity of the two data degrades, and the correlation value Zk is between 0 and 1 indicating the degree of correlation between the two data.

The calculation of the correlation value Zk is performed by k = Tm which can be extracted.
/ Td (steps S8 and S9). In addition,
The time width Td for shifting each of the reference data Yi, k corresponds to the detection accuracy of the delay time, and the required accuracy is usually ± 0.5 m.
Td is set to about 0.5 msec in accordance with the second.

Next, a search is made for the largest one of the obtained correlation values Zk (step S10). At the bottom of FIG.
6 conceptually illustrates the characteristics of the obtained correlation value Zk. Here, the delay time is calculated from the maximum correlation value Zp. However, if the principle is always followed, there is a possibility that an erroneous detection may occur. That is, although relatively similar to the reference data Xi,
There is a possibility that a part of the waveform that is originally unrelated may be erroneously searched. Therefore, it is determined whether or not the obtained maximum correlation value Zp is equal to or higher than a certain correlation estimable level S (step S1).
1) If this is not cleared (N in step S11)
O), returning to step S2, and re-executing the process from signal fetching. From a number of experimental results, this correlation estimation possible level S
It was found that setting the value to about 0.5 was appropriate.

If the maximum correlation value Zp is equal to or larger than S (YES in step S11), a delay time ΔT is calculated by the following equation (step S12). ΔT = Td · (p−1)

That is, the delay time ΔT is obtained as a value that is an integral multiple of the time width Td. Therefore, by setting the time width Td to be small, the detection accuracy of the delay time ΔT can be improved. Finally, the calculated delay time ΔT is output to the output unit 6 (step S13), and a series of processing ends.

FIG. 5 conceptually shows the waveforms and characteristics of the signals X and Y and the correlation value Zk in order to facilitate understanding of the detection operation described above. FIG. 1A shows signals X,
When Y is a simple single-pulse waveform, the correlation value Zk is
The maximum value clearly appears at a position corresponding to the delay time of both signals. FIG. 2B shows a case where the signals X and Y have AC components for several cycles, and the correlation value Zk has a characteristic in which the position corresponding to the delay time has a maximum value and droops before and after the maximum value. .

FIGS. 3 (4) show the case where a signal which is not the signal X which is intentionally delayed is set as the signal Y and is similarly input to the detecting device. Then, the correlation value Zk has the maximum value in two places (Zk ≧ S =
0.5), which indicates that there is a possibility of erroneous detection. On the other hand, FIG. 4D shows the input signals X, Y
Is the same as that of FIG. 3C, except that the reference time Tr is set longer. In this case, since the correlation value Zk is reduced and remains at a value sufficiently lower than the correlation estimable level S = 0.5, erroneous detection is avoided. As described above, by setting the reference time Tr to be long, the possibility of erroneous detection can be reduced.

By using the present apparatus, the performance of the conventional audio delay countermeasure monitor can be greatly improved. That is, when the delay time ΔT is obtained by the detection device, a signal XD is generated by delaying the signal X, which is an audio signal from the relay site, by the delay time, and the signal XD is converted from the signal Y, which is an on-air signal. By subtraction, the monitor signal YM is obtained. In this case, since the actual accurate value is used as the delay time, the monitor signal is
Only the sound of the local reporter is reliably removed, and the above-described adverse effects of the related art are surely eliminated.

Further, if the detection of the delay time is executed at regular intervals, and the subsequent arithmetic processing is performed using the value of the delay time sequentially detected, for example, a multi-relay program of three or more elements can be obtained. As described above, a desired monitor signal can always be obtained even in a case where the signal passes through the line differently, and therefore, the signal delay time is sequentially switched to different relay destinations.

Although the above description does not refer to a specific configuration method of the apparatus, by configuring a main part of the apparatus with, for example, a DSP (Digital Signal Processor), it is possible to perform calculations necessary for detecting a delay time. Processing can be performed at extremely high speed and accurately, and the hardware can be made extremely compact.

In the above description of the flowchart, in order to obtain the reference data Yi, k from the signal Y, the data Yj is once taken from the signal Y, and the reference data Yi, k is extracted from the data Yj. However, the reference data Yi, k may be directly captured from the signal Y by sequentially shifting the capture start time by the time width Td.

Further, the degree of correlation between the reference data Xi and Yi, k does not necessarily need to be evaluated using the correlation value Zk based on the above-described formula, and the correlation degrees of a plurality of combinations are relatively compared. Naturally, other formulas may be used as long as they can be performed.

[0036]

As described above, in the signal delay time detecting device according to the first aspect of the present invention, the signal X relating to the sound detected at the site of the sound generation and the signal containing the signal X are relayed. from the signal Y received monitor what was transmitted from the broadcast station via captures Xi and Yj at the sampling period Ts, further, from the latter extracts the reference data Yi, k by shifting time by one width Td, Both reference data Xi and Yi,
The delay time ΔT is obtained based on the maximum value Zp searched from the correlation value Zk with k, and the reference time Tr is set to 0.5 sec or more, and the calculation is performed when the searched maximum value Zp is 0.5 or more. Since the delay time ΔT is made effective,
The delay time can be easily and reliably detected from only the signals themselves before and after the delay occurs, without requiring complicated processing for adding a special signal.

Further, in the signal delay time detecting device according to the second aspect , once all of the fetched data Xi and Yj are below the predetermined noise level, the process returns to the data fetching process at that point and executes again. Therefore, it is possible to avoid a useless operation processing operation based on improper data fetch timing.

Further, in the signal delay time detecting device according to the third aspect , the maximum value Zp of the searched correlation value is 0.5.
Only in the above case, the delay time ΔT is calculated, so that the possibility of an output operation due to erroneous detection can be reduced.

Further, in the audio delay countermeasure monitoring apparatus according to the fourth aspect, the signal X is generated by delaying the signal X by the delay time detected by the signal delay time detecting apparatus , and the signal XD is subtracted from the signal Y. Monitor signal Y
Since M is created, a monitor signal YM from which the audio component of the signal X has been accurately and reliably removed can be obtained.

Further, in the voice delay countermeasure monitoring device according to the fifth aspect , the delay time is detected at predetermined time intervals and the value is updated, so that the delay time of both signals changes with time. In this case, a desired monitor signal is always obtained in response.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a signal delay time detection device according to an embodiment of the present invention.

FIG. 2 is a diagram conceptually showing the operation principle of the present invention.

FIG. 3 is a flowchart showing the operation of the present invention.

FIG. 4 is a flowchart following FIG. 3;

FIG. 5 is a diagram conceptually showing an operation example of the present invention.

[Explanation of symbols]

3 Central processing unit, 4 input units, 5 storage units, 6 output units, Xi fetched data (reference data), Yj fetched data, Yi, k reference data, Tm maximum delay time prediction,
Tr reference time, Ts sampling cycle, time width corresponding to Td detection accuracy, ΔT delay time.

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-48-95154 (JP, A) JP-A-3-173229 (JP, A) JP-A-6-13940 (JP, A) JP-A-4- 100424 (JP, A) JP-A-49-108913 (JP, A) JP-A-58-46735 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04H 7/00 G04F 13 / 00 H04B 17/00

Claims (5)

(57) [Claims] 1. A signal including a field signal according to the detected speech by X voice generating the signal X is input to the signal Y received monitor what was transmitted from the broadcast station via a relay means or the like , Removing the audio component of the signal X contained in the signal Y
In the signal delay time detecting device for detecting the delay time, data over a predetermined reference time Tr starting from a predetermined time (hereinafter, referred to as time t = 0) is taken in with respect to the signal X at a sampling period Ts. Captured data Xi (i =
1 to (Tr / Ts)) for the signal Y, the time (Tm + T) obtained by adding the reference time Tr to a predetermined delay time prediction maximum value Tm starting from the time t = 0.
r), means for taking in data at the sampling period Ts and storing it as taken-in data Yj (j = 1 to (Tm + Tr) / Ts), and setting a time width corresponding to the detection accuracy to Td (Td <
Tr), the start point is at time t = 0, Td, 2Td,.
.. Are sequentially shifted from each other and the data over the reference time Tr is extracted from the fetched data Yj, and the reference data Yi, k is extracted.
Means for storing as {i = 1 to (Tr / Ts), k = 1 to (Tm / Td)}, the above-mentioned fetched data Xi as reference data Xi, and expressed by the following formula from both reference data Xi and Yi, k Means for calculating a correlation value Zk (k = 1 to (Tm / Td))
The largest of the correlation values Zk, Zp = Zk (k = p)
Means for retrieving the value of k,
Means for calculating the delay time ΔT represented by the following equation from the value p of the following equation, and the reference time Tr is set to 0.5 sec or more, and the maximum value Zp of the searched correlation value is 0.5
A signal delay time detecting device wherein the calculated delay time ΔT is validated at the time described above. (Equation 1) 2. It is determined whether all of the fetched data Xi fetched from the signal X or the fetched data Yj fetched from the signal Y is at or below a predetermined noise level. 2. The signal delay time detecting device according to claim 1, wherein the data is rewritten by taking in from the signal Y and re-executing the subsequent operation. 3. The maximum correlation value Zp of the searched correlation value is 0.5.
3. The signal delay time detecting device according to claim 1, wherein when the value is less than the predetermined value, the signal X and the signal Y are taken in again, the data is rewritten, and the subsequent operation is executed again. 4. A signal delay time detecting apparatus according to any one of 請 Motomeko 1 to claim 3, said signal delay time signal X
Signal XD delayed by the delay time detected by the detection device
Means for creating a, Oyo by subtracting the signal XD from bicine No. Y, sound delay measures monitor apparatus characterized by comprising means for creating a monitor signal YM removing the audio component of the signal X. 5. The signal delay time detecting device detects a delay time at predetermined time intervals, and sequentially corrects a delay time for generating a signal XD based on the detected value. The audio delay countermeasure monitor device described in the above.