FR2680440A1 - Method and device for recording and replaying a subtitle on a video tape recorder and recording thus obtained - Google Patents

Abstract

The method makes it possible to record a program associating images with digital data on a tape, a cassette or video disc. After having extracted the digital data from the teletext of the incident video signal, in the format of a HD Val binary train, in which the data are formed in bytes, the packets constituting the subtitling magazine are extracted and they are accumulated in a buffer memory, the data sampled are serialised in the buffer memory at the rate of a local clock, which is slow compared to the frequency of the data bits in the video signal, the data are coded in manchester II code, and the signal obtained is filtered to attenuate the steep transition fronts, the packets, coded and filtered, are inserted into the active part of the video lines of an insertion window lying within the frame suppression interval, by amplitude modulating the luminance signal.

Description

METHOD AND DEVICE FOR RECORDING AND RETURNING
SUBTITLE ON MAGNETOSCOPE AND RECORDING THUS OBTAINED
The invention relates to the recording, on a medium such as a tape or cassette for a video recorder, of a program associating television type images with digital data representing subtitles with a view to its reproduction. The invention more particularly aims to make it possible to produce and use cassettes or tapes for video recorders comprising a subtitled television program, possibly in several languages, by implementing a digital teletitling or "teletext" system. not .

 It is not possible to record the multiplexed digital data to the video signal broadcast using a consumer video recorder, whichever of the current standards is used (Antiope, Ceefax, NABTS) because the bit frequency , 6 Mbits per second in the case of Antiope, is incompatible with the bandwidth of consumer video recorders which, moreover, provide a picture tainted by a significant jitter.

 One solution, described in patent FR 2 586 879, to which reference may be made, consists in remodulating the digital data in a form compatible with the recording on one of the sound tracks of a stereophonic VCR for the general public. This solution, sufficient for low bit rates, has the drawback of only applying to stereophonic video recorders and of mobilizing one of the two sound tracks.

 The invention aims to make possible the recording and playback of subtitled video programs using video recorders having a bandwidth significantly lower than the bit frequency of teletext standards, whether these video recorders are stereophonic or not. 'whether or not they are provided to record the blanking signal.

 It also aims to achieve this result without altering or reducing other functions of the video recorder.

 For this, the invention starts from various observations.

 The first is that a significant fraction of the information occupying the frame suppression periods (lines 16 to 22 of the odd frames and 329 to 335 of the even frames) does not relate to subtitling or is simply intended to protect the information by introducing redundancies.

This fraction is unnecessary or unnecessary for VCR recording and playback. It is therefore possible in particular to discard the broadcast control signals without inconvenience and distribute the useful data over all of the available lines and reduce the bit rate to a value compatible with the bandwidth of a consumer video recorder.

 Another observation is that the teletext decoders, whether they are connectable to the television receivers or whether they are incorporated in them, generally allow access to synchronization signals, of slow switching (switching from recording mode to playback mode). and fast switching and R, G, B components of the video signal.

The invention therefore proposes a recording method capable of being implemented by adding only a small number of components to a conventional teletext decoder, according to which, after extracting the teletext data from the incident video signal, under form of an HD Val bit stream in which the data is formatted in bytes,
- the packets constituting the subtitling magazine are extracted and accumulated in a buffer memory,
the data taken from the buffer memory are serialized at the rate of a slow local clock compared to the bit frequency of the data in the video signal,
- we encode the data in manchester II and we filter the signal obtained to attenuate the steep transition fronts,
- the packets, coded and filtered, are inserted into the active part of the video lines of an insertion window belonging to the blanking interval, by modulating the luminance signal in amplitude.

 The advantage of Manchester coding is that it carries in itself its sampling clock, since each transmitted binary element is represented by an upward or downward transition. Filtering, generally by an RC cell, makes it possible to reduce the bandwidth enough for the signal to be accepted by the video amplifier stages of a consumer video recorder.

 Practically Manchester encoding at a clock frequency of 526.6 kHz will generally be a satisfactory compromise between robustness and bit rate. The corresponding period is greater than the values for which there is a phenomenon of "magnetic hole in the recording tapes.

 The active part of a television line, following the line deletion signal, has a duration of 52 es: it is possible to transmit a bit rate of 15 Kbits per second, which is sufficient for subtitling in several languages.

 The insertion of data in the form of packets coded in Manchester, in the active part of the video lines of the insertion window implies a prior processing of these lines, constituted by a black alignment of the incoming video, an extraction of the signals of line and frame synchronization, counting of video lines with periodic reset (for example at each frame). An analog multiplexer then inserts the data into the luminance part.

 Typically, the insert window will occupy lines 16 through 22 of the odd frames and 329 through 335 of the even frames. A line, for example line 16, will generally be reserved for inserting a service package comprising in particular a datum identifying the magazine, and the content of the counter.

 The method can easily be adapted to the recording of subtitles provided by a local composition source.

 Upon return, the dual operations of the previous operations are carried out.

 The method according to the invention can be implemented using a multi-standard device, that is to say allowing for example to transcode both subtitling data conforming to the DIDON-ANTIOPE standard as to the CEEFAX standard, which differs in particular in that the CEEFAX data are always in the same location in the line. In all cases, it can also be noted that the method according to the invention does not prevent direct observation, on the television screen, of the program being recorded on a video recorder, with or without the subtitles.

The invention will be better understood on reading the following description of a particular embodiment of the invention, given by way of non-limiting example. The description refers to the accompanying drawings, in which
- Figure 1 is a block diagram showing how a device according to the invention can be incorporated into an assembly comprising a television, a teletext decoder and a consumer video recorder to allow the recording of subtitles and their reproduction;
- Figure 2 is a diagram showing a possible constitution of a transcoding device usable in an assembly of the kind shown in Figure 1, the links and the components involved in the recording of subtitles being shown in thick lines
- Figure 3 is a block diagram showing the operations involved in the recording in a device of the kind shown in Figure 2;
- Figure 4, similar to Figure 2, shows in thick lines the links and the components involved in the reading.

 The device which will now be described by way of example constitutes a unit intended to be connected to a television set, to a video recorder and to a decoder provided at least for receiving and viewing data conforming to the DIDON-ANTIOPE specification, as defined in the specifications disseminated by Télédifusion de France, to which reference may be made, all of these components being fitted with a SCART socket. However, a device according to the invention can also be incorporated in the decoder itself.

Figure 1 shows the TV receiver 10 with a SCART socket 12 (which could be replaced by another device with a tuner
VHF / UHF, capable of supplying and accepting the composite video signal in baseband) and a video recorder 14, also provided with a scart socket 16. This video recorder can correspond to any standard on the market but it must he too can supply and accept the composite video signal in baseband, view the subtitles and provide the digital multiplex DIDON on a scart socket 20. The decoder 18 can be of the Decan 2002 type, intended to decode the magazines Didon- Antiope and to be connected to a TV and a VCR by their SCART sockets.

 The device 22 according to the invention is provided with four interfaces constituted by three scart sockets 24, 26 and 28 and by an HDV socket 30 intended for carrying the digital multiplex DIDON accompanied by a clock H and the validation signal V. The socket 24 makes it possible to recover the video signals containing the data and the sound coming from the television set and to restore video and sound signals coming from the video recorder 14 as well as the RGB components coming from the decoder 18. The socket 26 is intended to provide the video recorder 14 with the video signals incorporating data in Manchester and sound and vice versa, to recover these signals when playing.

The video recorder also provides, via sockets 16 and 26, a slow switching signal
CL indicating its operating mode (reading or recording).

 The socket 28 supplies the decoder 18 with the video signal comprising the data coming from the television set 10 and recovers the synchronization, slow and fast switching signals and the RGB components coming from the decoder. The exchanges of information during registration are shown diagrammatically in FIG. 1 by solid lines, while exchanges during other operations are indicated by dashed lines.

 The power supply of the device 22 is advantageously provided from the decoder 18, by a link not shown: this is generally possible because the consumption of the device 22 is always low, 750 mA at 5 volts in a representative example.

 FIG. 2 shows only the components of the decoder 18 which are involved in the extraction and reinsertion of the subtitling data, that is to say the demodulator DIDON 32, the demultiplexer DIDON 34 and the display processor ANTIOPE 36 , containing a generator of characters to insert into the image.

 The device 22 is managed by a microcontroller 37, for example belonging to the family 8 031, provided with a program memory, a random access memory and an address decoding circuit 39. These elements do not ensure that the management of the processing circuits.

We will now describe the circuits involved in video processing. In Figure 2, and in the text that follows, the following notations are used
- SOUND out: outgoing sound to the TV
- SOUND in: incoming sound from the TV
- Y out: video signal leaving the TV
- Y in: video signal entering the TV
- CL: slow switching signal
- CR: fast switching signal
- R, G, B: red, green and blue components
- HDV out: clock, data and signal
validation leaving the decoder towards the
device;
- HDV in: HDV signal entering the decoder.

 During recording operation, the video recorder provides, via socket 26, a CL signal at a high level (usually 12 volts). The HDV out signals coming from the decoder 18 drive a DIDON 38 demultiplexer intended to extract the data from the subtitling magazine.

This demultiplexer is advantageously constituted by a CIDAC SA 5 250 circuit from the company RTC, which is multistandard, authorizes the processing of data in modes
CEEFAX, NABTS and DIDON and takes into account the various packet lengths possible in DIDON. The demultiplexer 38 includes a buffer memory organized in a queue, the minimum capacity of which is 2 kbytes, input and output interfaces and a sequencing controller.

 The demultiplexer 38 requires, for its operation, a sampling clock which can be supplied by the microcontroller or else taken from the signal H and which is regenerated internally by a phase locked loop.

The bytes representing the teletext magazine to be displayed (magazine 96) are stored in the RAM and read by the microcontroller, at a frequency of 526.6 KHz obtained by dividing the frequency of a quartz equipping the microcontroller or developed from H. The bytes thus read are transferred to a transcoding circuit
Manchester 40, which can be constituted by a gate or exclusive (XOR) between the data read in the memory of the demultiplexer and the sampling clock. In order for the steep-fronted Manchester encoded signal to be accepted by the video amplifier stages of the VCR, the signal is applied to an active low-pass filter 42 having a cutoff frequency of about 527 KHz.

 The bytes thus obtained must be grouped into packets and inserted into the video by a multiplexer 46.

For this, a serialization circuit 44 interposed between the random access memory of the demultiplexer 38 and the transcoder 40 can be controlled by a programmed logic network (not shown) which counts the bits in series and sends to the microcontroller a signal indicating that 'it must recharge the serialization circuit 44 at the end of each byte.

 The role of the serialization circuit 44 is to convert the bytes supplied by the microcontroller 37 into a serial binary train at the frequency compatible with the Didon demultiplexer 34 of the decoder 18. It is possible to use a microcontroller 37 which is not fast enough to supply the serialization circuit 44, which must receive a byte every 1.2 micro-seconds in the case of a Didon 34 demultiplexer at 6.2 Mbits per second by providing a buffer memory. This memory is loaded by the microcontroller 37. Then the microcontroller controls the serialization, outside the frame synchronization periods. Simple additional means, not shown, are sufficient for the serialization to start after a line synchronization.

 With a view to analog multiplexing by the video multiplexer 46, the packets coded in Manchester II and the incoming video are subjected to a processing making it possible to ensure the coherence of the signal. The Y out video signal from the TV is aligned to the black level by clamping to provide a level reference. Line and frame synchronization signals are extracted by a circuit 48 (LM 1881 for example) and used by an insertion window generator 50. The latter counts the lines in each frame and can be constituted by a programmable logic network providing a validation signal for the duration of the luminance parts of lines 16 to 22 and 329 to 335.

The device 22 is advantageously provided so that the data to be recorded can come from a source in the teletext language applied to an asynchronous serial interface 52, instead of coming from the captioning magazine extracted from the video (Didon channel in the case considered here ) supplied by a television receiver 10. I1 then comprises a data reception buffer memory and an input switch of the transcoder 40. This switch is automatically oriented on the channel which receives data, the other interface then being inhibited for a fixed period. Packetization mode with coding in
Manchester It depends on the source of the data. In the normal case of reception by the Didon channel, the processing begins as soon as a packet is received. In the case of reception by the serial interface 52, it begins after delay or as soon as the buffer memory (having a capacity corresponding to the maximum format of a Didon package) is met.

 The sequence of operations is thus, during recording, that shown diagrammatically in FIG. 3.

 We will now describe, by way of example, closed caption data signal structures which have been shown to be satisfactory for recording and playback, but are not exclusive.

Relationship between data signal and image signal
Advantageously, the A / S ratio is given a value of the order of 7/3, A designating the nominal amplitude of the waveform constituting the data signal, with a duration of 52 rs per line, and S designating the nominal amplitude of the waveform constituting the line synchronization signal. This last signal has a duration of about 10.5 ps, between the mid-amplitude point and the first transition from the low state to the high state of the data signal. The line synchronization signal retains its usual form, consisting of a transition to the low level followed by a plateau at the suppression level.

Package formats
The packets transmitted to the video multiplexer 46 by the transcoding circuit 40 all have the same structure and the same length, but they can be of several different types.

A satisfactory structure consists of synchronous line synchronization packets, each having 27 bits, consisting of
- 2 SYN synchronization bits
- 1 TPP bit indicating the type of packet (taking a value if the packet contains only useful data and another if it also contains other information)
- 3 bytes of data.

 The packet contained in line 16, sent to each image during the first frame, can have a particular function and contain the service data intended on the one hand for retrieving the images (by counting them) on the other hand to the configuration of the video recorder.

This line packet 16 can also include two SYN synchronization bits, 1 additional bit, then 3 bytes.
Image counting can be done using the first two bytes and the additional bit: by counting only even images, every 80 ms, the counting capacity of 131,071 is sufficient in most cases. However, it is possible to increase the maximum counting time by counting only one image among several.

The last byte of the line packet 16 can include 4 bits indicating the number of lines occupied by the packets in each frame of the video signal, that is to say the size of the data insertion window. The last 4 bits of the last byte can be used to indicate the type of original coding of the data (Didon,
Ceefax, TTY, etc ...) and the magazine that has been recorded (Teletext or subtitling).

 The first packet of each message or group of data can have a synchronization role and be indicated by a value of the TPP bit different from that of this bit in standard packets. The first byte of such a synchronizing packet includes a group continuity index (which, in the case of Didon, is the continuity index of the Didon packet without the redundancy bits) and a group format indicator, which n is of interest only in the case of variable lengths, which is that of Dido, but not of Ceefax.

 When playing with playback on a television receiver fitted with a scart socket having terminals R, G, B, the connections are those shown in solid lines in FIG. 4. These connections are established by coming to the low level of the slow switching signal CL from the video recorder. Certain assemblies have not been detailed in FIG. 4 for simplicity.

The operations carried out are then reciprocal from the previous ones. The video signal Yin supplied by the video recorder is analyzed. If it contains data coded in
Manchester II in the frame deletion interval, this data is extracted and processed.

In the particular case envisaged here, the data is converted to the Didon protocol and sent, via the interface 30, to the decoder 18. If the latter is in subtitling mode, that is to say if the magazine 96 has was selected in the case Dido, the visualization processor
Antiope 36 which it contains generates the characters and embeds them in the RGB signal which is sent back to the television receiver by interfaces 20, 28 and 24.

 More precisely, the signal Yin is sent to the output video multiplexer 54 of the device 22 on the one hand directly, on the other hand by a processing chain incorporating the decoder 18.

 In the processing chain, the video signal Yin is filtered to remove the chrominance subcarrier generated by the video recorder, then applied to the line and frame synchronization extraction circuit (not shown) and, via a packet extraction circuit 55, at Manchester 56 decoder. In circuit 55 the signal is aligned to black and then compared to a reference voltage during the frame return periods, for example in a LM 306 P comparator from National Semi Conductor .

The clock is extracted from the coded signal by a flip flop detecting the edges and it is used to sample the data. The same clock is delayed, then used to control a deserializer which reconstructs bytes. For each byte, an interrupt is sent to the microcontroller 37 to cause the reading.

 The packets thus processed are written to the input buffer memory of an HDV 58 adapter. Serialization is carried out by this adapter as soon as a packet is fully loaded. Finally, the HDV multiplexer 60, controlled by the slow switching signal CL, sends the packets in HDV format to the Didon demultiplexer 34 which in turn feeds the display processor 36.

 The signals R, G, B and the fast switching signal CR supplied by the Antiope display processor 36 are sent, by the interfaces 20, 28 and 24, to the television set and attack an overlay circuit 62 which also receives, by the by means of a multiplexer 64 controlled by the slow switching signal CL, the unprocessed video signal Y.

 As indicated above, the device can also be used to copy captions supplied by a video recorder using a mother tape and connected to the interface 24; when it is used in conjunction with a teletext decoder (or incorporated in such a decoder), it also makes it possible to restore a video signal with embedded subtitles to a television receiver or monitor. It also makes it possible to reproduce a program recorded in house on cassette or tape, having the same characteristics as those which it gives to the recording itself. Consequently, the subject of the invention is also a recording medium usable on a consumer video tape recorder, constituted by a recorded tape or cassette, making it possible to restore to a television receiver or monitor a video signal with embedded subtitles. at least one video track and one sound track, the video track of which contains the recording of the image and the recording of the subtitles. Subtitles, excluding other teletext magazines, are recorded in Manchester II encoding, at a clock frequency of approximately 527 kHz, in the form of packets occupying each line of a window in the frame blanking interval of each frame, said packets each comprising at least one line synchronization header and three bytes representing the subtitle data, the first packet of each data group incorporating a continuity index recorded without bit redundancy.

 From a first recorded support, it is without difficulty, by the usual reproduction methods, to make replicas on a support of the same kind or a different support. It is in particular possible to transfer the content of a magnetic medium (tape or cassette) to a videodisc for playback using a player other than a video recorder.

Claims (9)

 1. A method of recording, on a medium such as a tape or cassette for a video recorder, a program associating television images with digital data representing subtitles, with a view to its reproduction, characterized in that, after extracting the digital teletext data from the incident video signal, in the form of an HD Val bit stream in which the data is formed in bytes,  - the packets constituting the subtitling magazine are extracted and accumulated in a buffer memory,  the data taken from the buffer memory are serialized at the rate of a slow local clock compared to the bit frequency of the data in the video signal,  - we encode the data in manchester II and we filter the signal obtained to attenuate the steep transition fronts,  - the packets, coded and filtered, are inserted into the active part of the video lines of an insertion window belonging to the blanking interval, by modulating the luminance signal in amplitude.  2. Method according to claim 1, characterized in that a Manchester II coding is used at a clock frequency of 526.6 KHz.  3. Method according to claim 1 and 2, characterized in that the data encoded in Manchester are grouped in packets each occupying the useful part of a line, each having three bytes of data and a synchronization header.  4. Method according to claim 3, characterized in that the first packet of each group of data contains a group continuity index and a group format indicator deprived of redundancy bits.  5. Method according to claim 3 or 4, characterized in that one inserts in line 16 a packet containing an associated image number and possibly the size of the insertion window.  6. Method according to any one of the preceding claims, characterized in that the report is given A / S a value of the order of 7/3, A designating the nominal amplitude of the waveform constituting the data signal, with a duration of 52 ps per line, and S designating the nominal amplitude of the waveform constituting the line synchronization signal.  7. Device for recording, on a medium such as a tape or cassette for a video recorder, and for restoring a program associating television images with digital data representing subtitles, comprising a recording part, a restoring part and connection interfaces with a video recorder, a television receiver and a teletext decoder, characterized in that  - the recording part comprises a demultiplexer (38) for extracting the subtitle magazine data supplied by the decoder (18), having a random access memory, a Manchester transcoding circuit (40) followed by an active low-pass filter (42), and a video multiplexer (46) which is controlled by a synchronization extraction circuit (48) for inserting the coded packets Manchester II and filtered in the active part of the video lines of a window which belongs to the blanking interval, by amplitude modulating the luminance signal,  and in that the restitution part comprises a circuit for extracting packets from the video signal supplied by a video recorder by the corresponding interface, a Manchester decoder (56), and an HDV adaptation circuit intended to supply the decoder of teletext (18).  8. Recorded medium for consumer video recorder, consisting of a tape or cassette with at least one video track and one sound track, the video track of which contains the recording of the image and the recording of the subtitles, excluding other teletext magazines, recorded in Manchester II coding, at a clock frequency of about 527 kHz, in the form of packets occupying each of the lines of a window in the frame blanking interval of each image, said packets comprising each at least one header for synchronization with the line and three bytes representing the subtitling data, the first packet of each group of data incorporating a continuity index recorded without redundancy bit.  9. Recorded medium consisting of a video disc whose video track contains the recording of the image and the recording of the subtitles, excluding other teletext magazines, recorded in Manchester II coding, at a frequency of clock of about 527 kHz, in the form of packets occupying each of the lines of a window in the frame blanking interval of each image, said packets each comprising at least one header for synchronization with the line and three bytes representing the captioning data, the first packet in each data group incorporating a recorded continuity index without redundancy bit.