JP2624538B2 - Audio synchronization method for television format conversion - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to television format conversion, and more particularly, to lip synchronization (synchronization) of audio with respect to converted video.

SUMMARY OF THE INVENTION The present invention relates to a conversion method between two television systems having slightly different field frequencies from each other, and a memory device capable of independently writing and reading data having a capacity exceeding one frame of video. Use 2
Along with video signal processing for appropriately deleting or inserting a screen corresponding to a difference in field frequency between two television systems, a buffer memory capable of performing writing and reading independently for an audio signal is also provided. hand,
A / D, D / A or D provided on the input / output side of the buffer memory
The ratio of the sampling frequency of the / D converter is changed appropriately and the lip sync with the video in the previous video conversion is obtained.

By doing so, it is possible to follow the conversion of video between two television systems whose field frequencies are slightly different from each other, and lip sync between video and audio without unnaturalness can be obtained.

(Conventional technology) A method of converting between two television systems having slightly different field frequencies is disclosed in
No. 104866 "Television system conversion method" is available, and it is possible to perform system conversion on video without any problem.

FIG. 2 employs a configuration diagram of an example of a system conversion apparatus configuration described in the specification of the above-mentioned Japanese Patent Application Laid-Open Publication No. Sho.
The figure shows a case where TV (television) system 1125 lines / 60 fields are converted into NTSC system 525 lines / 59.94 fields.

An outline of the operation will be described with reference to FIG.
HDTV 1125/60 video signal is first converted into a field conversion circuit 1.
In step 5, the number of scanning lines is converted into a 525/60 video signal by a known technique. The converted signal is temporarily written to a buffer memory 16 having a memory capacity of one frame or more, and then converted into an NTSC 525 / 59.94 video signal by reading independent of writing. The control is performed by a write address control circuit 19 and a read address control circuit 20 in FIG.

The problem with this configuration is that 60 fields of video input, 5
Since signals are gradually accumulated in the memory for outputting a 9.94-field video, an extra signal must be deleted somewhere in units of one frame. In this embodiment, this operation is executed by the scene change detection circuit 17 and the motion detection circuit 18 by capturing the time of a scene change or the time of a still image.

Although a more detailed description is given in the above-mentioned Japanese Patent Application Laid-Open, the purpose of video conversion has been sufficiently achieved in any case. However, only the delay of the buffer memory (about one frame) is added to the sound.

As a conventional example of only an audio signal, as a method of synchronizing audios having slightly different sampling frequencies, a method using a variable buffer and silent interval detection is disclosed in US Pat. No. 4,716,472, "Variable speed replay of digital audio".
dio with constant output sampling rate ", but this patent does not mention lip sync with video.

(Problems to be Solved by the Invention) As described in the section of the prior art, regarding the conversion between two television systems having slightly different field frequencies, the video signal is converted to the one described in JP-A-59-104866. If the conversion is performed according to the method, the system can be converted without any problem. However, no method has yet been proposed for the audio signal that can satisfactorily obtain the converted video and lip sync.

Therefore, an object of the present invention is to provide an audio synchronizing method that accompanies a television system conversion in which the converted image can be lip-synced with the converted system and can perform system conversion of an audio signal without unnaturalness. is there.

(Means for Solving the Problems) In order to achieve this object, the audio synchronizing method according to the television system conversion of the present invention converts the system conversion between two television systems whose vertical scanning frequencies are slightly different from each other by a video signal. And a memory device capable of performing writing and reading with a capacity exceeding one frame independently of each other is provided, and a point in time when a screen corresponding to the difference in the scanning frequency between the two television systems is deleted or inserted is determined. A television set arbitrarily set within a range of a memory capacity of a memory device, and deleting or inserting the screen when a correlation value of an image between adjacent frames or fields is out of a predetermined range. In the system conversion system, writing and reading for audio signals can be performed independently of each other. Buffer memory, and an input / output side of the A / D converter having a variable sampling frequency in order to make the writing and reading speeds variable, a D / A converter or a D / D converter. When the screen is deleted or inserted, the ratio of the sampling frequency of the input / output converter is changed correspondingly to maintain the lip sync.

(Example) Hereinafter, the method of the present invention will be described in detail with reference to the accompanying drawings by way of examples.

FIG. 1 is a block diagram showing the configuration of an embodiment of an apparatus for carrying out the method of the present invention. The video processing portion is the same as the conventional configuration shown in FIG. First
The illustrated video buffer memory 2 corresponds to the buffer memory 16 shown in FIG. 2, and the input / output side thereof in FIG.
An A / D converter 1 and a D / A converter 3 are provided. Also this A / D
The analog video input signal to the converter 1 is a signal obtained by converting the number of scanning lines of the HDTV 1125/60 to 525/60. Also the second
It is assumed that the illustrated scene change detection circuit 17, motion detection circuit 18, write address control circuit 19, read address control circuit 20, and the like are incorporated in the control circuit 5 shown in FIG. For audio signals, an audio buffer memory 10 and an A / D converter 9 and a D / A converter 11 are provided on the input and output sides, respectively.

The HDTV signal has a field frequency of 60.00Hz
It is prescribed. Therefore, in order to convert from HDTV to NTSC, it is necessary to convert the number of scanning lines and 1125 lines to 525 lines, and also to convert the field frequency from 60.00 Hz to 59.94 Hz.

To do this, the video signal arriving at 60.00 Hz is once written into the buffer memory 2 according to this synchronization, and
If this is read out according to the synchronization of z, it is possible to carry out system conversion. However, in this embodiment, since the writing speed is faster and the reading speed is slower, the amount of information stored in the buffer memory 2 gradually increases. Overflows. For this purpose, the time point of the rest of the screen and the time point of the scene change are found, and the information in the buffer is reduced by deleting the frame.

This is done about once every 33 seconds on average. If the material to be converted contains many scene change points, still images, etc., the operation is to delete once every 33 seconds on average, and there is no problem. However, when the material is a material such as a moving game using only one camera, the material hardly includes a still image or a scene change point, and the buffer memory is easy to use. Overflows. At this time,
Even if motion is included in the image, the frame must be forcibly deleted. In the converted image, the motion is traversed, which hinders the motion.

Therefore, it is conceivable to increase the capacity of the buffer memory 2. At this time, if a scene change point or the like does not appear, image information is stored in the buffer memory 2 steadily, and a delay occurs in an output image. At this time, if only attention is paid to the video, no unnatural feeling will occur at all, but if no countermeasure is taken except for adding the delay amount corresponding to the memory capacity described in the section of the prior art to the audio, Since this advances relatively to the image, it gives a very strange feeling. The present invention seeks to improve this point.

That is, the audio signal is converted into a digital signal by the A / D converter 9 in the same manner as the video, and then the audio buffer memory 10 is used.
Once. Then, the data is converted again by the readout D / A converter 11 to be converted into an analog output, or is directly connected to a digital VTR signal input. Here, it is assumed that the read clock is constant, but the same object can be achieved by slightly changing the clock on the D / A conversion side.

For audio, the sampling frequency for A / D and the sampling frequency for D / A are basically the same. In this case, the sampling clock for A / D conversion is about 1/1000 (6
A / D conversion is performed for a while after increasing the ratio (because the ratio of 0.00Hz and 59.94Hz is 1001: 1000). In this case, the number of samples of the audio signal increases, so that the audio information stored in the audio buffer memory increases. That is, a delay occurs in the audio signal output after the D / A conversion (the change in the pitch is less than the detection limit and does not give an unnatural feeling). In this way, the audio signal can be delayed by an amount corresponding to the delayed video signal, and as a result, lip sync can be achieved.

Even if there is no point in the image signal where frames can be deleted such as a scene change point for several minutes, it is expected that a point in time where frames can be deleted comes later. At this time, if the frame is deleted, the image information is advanced by one frame, and as a result, the audio signal is relatively too late. Therefore, when this condition is satisfied,
A / D converter clock frequency approximately 1/1000 compared to standard time
Lower. Then, the number of audio information samples decreases, and the amount of information in the buffer 10 decreases. That is, the sound is relatively advanced, which makes it possible to take lip sync again.

The lip sync can be obtained by slightly changing the sampling frequency applied to the audio A / D converter 9 as described above. However, if a sudden change is made when the clock is changed, the audio to be finally output is output. Since there is a possibility that the change of the clock may be detected, it is desirable to change the clock by taking about 1 to 5 seconds.
The control required for these operations is performed by the control circuit 5.

Conversely, when converting from 59.94 Hz to 60.00 Hz, regarding the video, information is first put in advance until the buffer is full, and then the video is read out in synchronization with 60.00 Hz. In this case, it is necessary to increase the number of frames by compensating for the (frame) by finding the time point of the still image and repeating the frame. However, if the timing for this purpose does not arrive at all, the information stored in the memory is gradually increased. Remaining low. That is, the video advances with respect to the audio, and the lip sync deviates. At this time, the control circuit 5 similarly controls the audio A / D converter 9 or D / A
By controlling the sampling clock of the converter 11, the delay of the audio signal corresponding to the delay amount of the video signal can be realized.

Also, the relative delay between audio and video is very noticeable when audio precedes. Also, if the sound is too late, the user will feel unnatural. Therefore, in this case, the advance of the voice is set as the limit of advance by one frame, and the delay of two frames is set as the limit of delay. If the sound is controlled to be within this limit, no practical problem will occur.

It is also possible to replace the A / D and D / A converters of this device with an all-digital system.

Although the embodiment of the method of the present invention has been described in detail above, the present invention is not limited to this embodiment, and changes within the scope defined in the claims of the present invention,
The correction is self-evident.

(Effects of the Invention) The audio synchronization method of the present invention has been described in detail above. However, according to the method of the present invention, the method conversion between two television systems having slightly different vertical scanning frequencies is one.
In a conversion method in which a video signal is format-converted through a memory having a capacity exceeding a frame amount and a frame change or an image frame is deleted or inserted using a time point of a video scene change or a still image, the video signal and the lip sync are also converted. Simultaneous system conversion of the taken audio signal and the video signal can be performed, and even if the capacity of the video memory is large, the lip sync is not lost and the unnaturalness that has conventionally occurred between the video and the audio can be removed.

[Brief description of the drawings]

1 and 2 show block diagrams of the configuration of an embodiment according to the present invention and a conventional example, respectively. 1,9 A / D converter 2 Video buffer memory 3,11 D / A converter 4,6 PLL circuit 5 Control circuit 7 Audio sampling frequency converter 8 ... Countdown circuit 10 ... Audio buffer memory 15 ... In-field conversion circuit 16 ... Buffer memory 17 ... Scene change detection circuit 18 ... Motion detection circuit 19 ... Write address control circuit 20 ... Read address control circuit

Claims (6)

(57) [Claims] The vertical scanning frequencies are slightly different from each other.
A memory device capable of independently performing writing and reading of a video signal with a capacity exceeding one frame for video signal conversion between two television systems is provided, and a difference in the scanning frequency between the two television systems is provided. The point in time when the corresponding screen is deleted or inserted is arbitrarily set within the range of the memory capacity of the memory device, and when the correlation value of the image between adjacent frames or fields deviates from a predetermined range, In a television system conversion system in which a screen is deleted or inserted, furthermore, a buffer memory capable of performing writing and reading for an audio signal independently of each other, and a variable writing and reading speed on its input / output side. A / D converter, D / A converter or D / D converter with variable sampling frequency A television converter, comprising one converter, and changing the sampling frequency ratio of the input / output converter correspondingly when the screen is deleted or inserted, so as to maintain lip sync. Voice synchronization method with. 2. A synchronization method according to claim 1, wherein a time when a correlation value of an image between said adjacent frames or fields is out of a predetermined range is a time when a scene change or a still image state is detected. An audio synchronizing method for television system conversion. 3. A synchronous method according to claim 1, wherein the sampling frequency of the converter at the output side of the buffer memory is fixed and the sampling frequency of the converter at the input side of the buffer memory is fixed when the ratio of the sampling frequency is changed. An audio synchronizing method for television system conversion, characterized by changing a sampling frequency of a converter. 4. The synchronization method according to claim 1, wherein, when changing the ratio of the sampling frequency, the change is determined by reciprocal of a frame vertical scanning frequency difference between the two slightly different television systems. A method for synchronizing audio according to a television system conversion, characterized in that the audio synchronization is performed gradually over a period of time equal to or less than half of the time represented. 5. The synchronization method according to claim 1, wherein the converter on the input side of the buffer memory is an A / D converter.
An audio synchronization method for television system conversion, wherein the converter on the output side is a D / A converter or a D / D converter. 6. The synchronization method according to claim 1, wherein the lip sync for the audio video is performed within one frame of the video when the audio signal advances with respect to the video signal. A method for synchronizing audio with a television system conversion, characterized in that when a signal is sent, it is within two frames of a video.