JPS6383970A - Data scrambling system for pcm multi-track system - Google Patents

Abstract

PURPOSE:To eliminate the ununiformity of the error rate to correct the burst error by preventing pairing channel data from being recorded on one track when a pair of data of each channel data and a pair of data of another channel are combined and distributed. CONSTITUTION:Each of areas into which the recording area of one rotation of a drum 9 is divided by N is used for one PCM code sequence, and information of each channel is divided into a pair of even data and odd data. When a pair of data of each channel and that of another channel are combined and are distributed to areas, recording of pairing data on one track 3 is inhibited. When one track is divided by (m), a distance D is set in every area in the tape breadthwise direction from the tape center and a sum Dt of distances of recording areas of pairing data is set to satisfy equations. Thus, respective channels are reproduced at an about certain error rate to eliminate defective channels.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a data scrambling method (interleave or dinterleave) in a PCM multi-track system.

[Conventional technology]

Helical scan recording using a rotating head is a method of recording diagonally to the magnetic tape as a recording medium, but when recording multi-channel information data at the same time, it is necessary to use a multi-head or multiple It is possible to make it into a truck. Considering the conditions for a recorder/player, it is desirable that information data be processed in real time and output as much as possible in order to save internal memory.Therefore, the cycle of information data to be PCM encoded is generally performed by rotating a drum or It takes the same value. In addition, encoding is determined based on the error rate under the conditions of the equipment, and how much correction ability is required.
Redundancy and correction processing time are also determined. Therefore, since the amount of information that can be processed by one decoder is determined, many decoders are required for simultaneous multi-channel playback. At this time, it is conceivable to divide and process the pattern recorded on the recording medium based on the PCM code sequence handled by one decoder.

Such a helical scan type multi-channel PCM
As a recording device, an 8x video format using an 8x*ti-track has been proposed. This recording pattern is shown in FIG.

In the figure, 1 is a recording tape, and 2 is a track scanned by a recording head, that is, a recording pattern. Each trunk 3, 3, . . . of each recording pattern 2 is divided into six recording areas a, to a6, each of which corresponds to an independent input signal. However, in one area, a "pair" of data L/R of the same type (L is a left signal, R is a right signal) coexists, resulting in 12 channels of data.

[Problem that the invention seeks to solve]

In this case, if a burst error occurs in area at, for example, in the tape longitudinal direction (tape running direction), it becomes impossible to correct the data in area a1.

Furthermore, even if a burst error occurs in the head scan direction, the erroneous data cannot be correctly decoded because the channel data is configured (completed) in this one area.

There are various possible causes of the above two burst errors, but the former burst error in the tape running direction can be caused by non-uniformity of the tape path, contact of the head with the tape, etc. Especially on the tape,
In the lower area, due to the nature of the rotary head, it is generally known that the error rate, that is, the rate of false occurrence, has a characteristic as shown in FIG. In the figure, the horizontal axis shows the position in the tape width direction, and the vertical axis shows the error rate.

As can be seen from FIG. 2, even in the best recording and reproducing conditions, the data recorded in area "3+84" near the tape center and the area "a" near the tape edge.

There is a basic difference from the data recorded on +86.

Also, the cause of the latter burst error in the head scanning direction is a problem with the amount of head protrusion.

It is possible that dust has adhered to the head.

Furthermore, when recording/reproducing multi-channel signals, the greater the number of channels, the more corresponding recording processing circuits and reproduction processing circuits are required. At this time,
For example, if we focus on the reproduction system, it is necessary to scramble (interleave) the data in order to create a recording pattern that is advantageous in terms of magnetic recording or mechanism. p handled in a container
There is a problem in that increasing the number of types of data or limiting the number of types of data may lead to non-uniformity in processing time.

The present invention compensates for the drawbacks of the prior art, provides the most ideal recording pattern in the case of burst errors in the tape running direction and the head scan direction, and furthermore makes it possible to equalize the data processing time during playback. It is. That is, the present invention proposes that, in this recording pattern, various errors that occur on magnetic recording be taken into account, and the errors should be scrambled so that they can be decoded by a correction circuit as much as possible.

[Means for solving problems]

The present invention has N (more than 3) information channels, M (≧2) tracks 3 and a plurality of heads 11, A, B,
In recording pattern 2 in a rotary head type PCM multi-track system having C, the area obtained by dividing the recording area of one rotation of the drum 9 into N is set as an I PCM code sequence, and the information of each channel is divided into two parts, one for EVEN data and one for ODD data. When dividing each channel data set into two sets forming a "pair" and combining each set of channel data with another set of channel data, and distributing the combination to the above area, the sets forming the "pair" is prohibited from being recorded, and when one track is further divided into m, from the tape center to the tape width direction,
This data scrambling method in a pcFI multi-track system is characterized in that a distance is set for each area, and the sum of the distances of the recording areas of the pair of recording areas is set as follows.

[For production]

In this way, data of each channel is converted to EVEN and 0DDO.
Dividing into sets and creating PCM code series data for each allows for intermediate value correction to be performed using the data in the other PCM code series even if an error occurs in that PCM area due to a burst, etc. It is possible.

Also, the sum D of each distance of the recording areas of the above “pair”
By setting t such that each channel is reproduced at a substantially constant error rate, it is possible to eliminate defective channels.

〔Example〕

The present invention is particularly effective in a system in which one track is divided into three or more areas and each area is treated as an IPCM code sequence block.

Next, referring to FIGS. 7 to 9, a case will be described as an embodiment of the present invention in which 12 channels of information data are divided into 12 PCM code sequences and recorded. As shown in Figure 7, the number of tracks 2 is 2, and 1 track and 7 tracks are divided into 6 areas al la a
z...Divide into a6. First, regarding the information data structure within the CM code sequence, as mentioned in the prior art section, one PCM code area causes a burst error for some reason, making it impossible to decode that part. With this in mind, we will create a mixed channel format. As mentioned above, by splitting and encoding one channel's data, even if the area where the burst error occurs becomes undecodable, it can be corrected by decoding in other areas where that channel data exists. It becomes possible.

In other words, the EV of input information data arrangement of one channel
By dividing by ENloDD, even if the PCM code sequence area where EVEN data exists becomes undecodable, O
If the area where the DD data exists can be completely decoded, all the EVEN data that has become decodable can be decoded by intermediate value correction. In other words, since EVEN data and ODD data are the same type of information, they cannot be paired with each other.
It is something that does. EVE each 12 channel data
The number of combinations when dividing into N data and ODD data sets and combining each PCM code sequence set again is 24C2, but here the I PCM area is always EVEN and ODD data.
If the combination of EVEN and 0DDO of two different channels is decided,
Same different channels 0I)D and EVENO
The sets shall be combined.

For example, if 1 channel of EVEN and 2 channels of ODD are combined, 1 channel of ODD and 2 channels of ODD are combined.
Channel EVEN shall also be combined.

Above, channel 1-EVEN, channel 2-
The combination of ODD and the combination of channel 1-〇〇〇 and channel 2-EVEN are also “pairs” with each other.

By forming "pairs" of combinations in this way, it is possible to reduce the number of information channels handled by one decoder.

Now, the combinations 1 to 6 are set, and the combinations 1' to 6' are set as their "pairs."

Now, regarding the recording positions of the 12 PCM code sequences 1 to 6 and 1' to 6', let's first focus on the track direction (HEAD 5 CAN direction). If a burst error occurs in this direction, the pair of bears mentioned above are located on the same track, the two channels of bare information become completely undecodable. Therefore, it is desirable to set the PCM areas so that a "pair" of PCM areas does not occur on the same track. Therefore, as shown in FIG. 8, each area al of one track 3...
Combinations 1 to 6 are recorded in a6, and combinations 1' to 6' are recorded in areas a1 to a of other trunks 3. As a result, there are no paired combinations on the same track, so even if a burst error occurs in one track 3, the data on that track can be decoded from the data on the other track 3. Also, a pair of combinations 1.
1', 2.2', . . . are not lined up in the running direction of the tape. As a result, even if a burst error occurs in the running direction of the tape, the data of the combination in which the error occurred can be decoded by the data of the other combination that forms a "pair" with it.

Figure 9 shows the case where the track is divided into 5 areas, and the "pair" combinations 1 to 5.1' to 5' are similarly recorded in areas a1...a. .

Next, the distance of each area in the tape width direction is defined as a setting condition. As shown in FIG. 7, each trunk 3 of the recording pattern 2 of the tape 1 has six areas a,
It is divided into...a.

The distance of each area al r ag... is the distance from the center area a4 indicated by diagonal lines, and the distance of the center area a4 is D=O.Area a
3°a4 distance D=1. Further, the distance D of area a2 + a is 2°, and the distance D of area a is 3.

Now, as mentioned in the conventional example, the error rate in the rotating head type playback system is as shown in Figure 2. Therefore, the comrades that form the "pair" mentioned above, for example, are set at the tape edge (top, bottom). If this were the case, the two channels in that group would always have poor quality playback. Therefore, in order to equalize this, the sum Dt of the distances of each pair of recording areas is determined as follows.

When configuring a PCM code sequence by dividing one trunk into m segments, when one set of recording areas is set, the sum of the distance of that area and each distance of the other "paired" set of recording areas is calculated. In the previous example, Dt = m/2 (m = even number) is set at a distance. (The distance here is not related to the longitudinal direction of the tape.) In the above example of 12 channels, 2 tracks, and m = 6. is shown in Figure 8.

Also, when m is divided as an odd number, it becomes 10
channel, m = 5 for 2 tracks (5 for 1 trunk)
FIG. 9 shows an example of division).

By doing this, the sum of the error rates in each pair of recording areas is equalized, so each channel is reproduced at a substantially constant error rate, and defective channels can be eliminated. It is something that can be done.

Hereinafter, another embodiment of the present invention will be described with reference to FIGS. 3 to 6, taking as an example the recording pattern of a rotary head type 12-channel multi-track recorder/player.

In Figure 3 (a), 9 is a drum and 10 is a tape.

11 is a recording head, and 12 is a guide pole.

In this case, three recording heads 11 from A to C are attached to the drum 9, and one revolution of the drum 9 (see Fig.
Record pattern 2, ie, three tracks 3.3.3, as shown in II+). There are 12 channels of recorded information, which must be stored within three 3°3.3 tracks. To divide trunk 3 into areas, the number of channels N=12. l - Number of racks M = 3, so M/N = 0.25 = 'A, one trunk 3 is divided into four as shown in Figure 4, and the area divided into four is al t ag + a,, +
The IPCM code sequence is recorded in 84. This means that the correction operation is completed with the code within this area.

On the other hand, since there are 12 recording information channels, the information of each channel is divided into even numbers, and the odd numbered (ODD) and even numbered (EVEN) are divided into - sets.
/EVEN to create each “pair”. Each channel set is then recombined, but the same channel sets or "pairs" are prohibited from being combined.

As an example, the above sets may be combined as follows.

1: 1 channel-EVEN, 7 channel/I/
-0DD2: 2 channels -EVEN, 8 channels/L/ -0DD3: 3 channels JL/-EVEN
, 9 channel-0DD4:4f+7 Nehru E
VE8.10 channel-ODD5:5 channel-
EvEN, 1iJ-+Cinelu0DD6:6f-
p7 Nehru EVEN, 12 channels-〇D01'=
7 channel EVEN, 1 channel-〇〇0
2': 8 channels-EVEN, 2 channels/L/
-0DD3': 9 channels -EVEN, 3 channels -0DD4': 10 channels 7 Nehru EVEN, 4
Channel-〇DD5': 11 channels/l/-EVE
N, 5 channels/Iz-ODD6': 12 channels + 7 channels/I/-EVEN, 6 channels〇〇〇 above 1
For two combinations 1 to 6.1' to 6', the number of channels in parentheses may be set arbitrarily, but they must be combinations other than ODD/EVEN of the same channel, that is, combinations other than "pairs". , satisfy the condition of setting "pairs" of combinations for the data of the combined channels (same number as in the above example, 6 pairs each with ′ and 6 pairs). It is something that exists.

In this way, dividing the data of each channel into pairs of EVEN and ODD and creating each PCM code sequence data means that the PCM area at (
For example, even if an error occurs due to a burst or the like in one channel (EVEN), the intermediate value can be corrected using the data (for example, one channel ODD) in the area 3 of the other first track. or"
Setting a pair of combinations means reducing the number of channels processed by one decoder.

FIG. 4 shows an example of the scrambling method of the present invention. Further, FIG. 5(A) shows the output timing of each reproducing head when this scrambling method is used, and FIG. 5(TI) shows the processing timing of the decoder at this time.

In FIG. 4, recording pattern 2 has a configuration in which one track 3 is divided into four as described above, and data for 12 channels is inserted into a total of 12 areas in the 12 combinations described above. . The characteristics of this pattern are as follows: 1) When focusing on a specific channel, the PCM code sequence containing the other data of that channel (ODD data in the case of EVEN data) is
2) The above-mentioned paired areas are located near the edge of the tape, and the other is set in the center of the tape, and 3) When viewed in the longitudinal direction of the tape. There are no pairs of tapes in the same longitudinal direction, and 4) PCM code sequence areas that form a pair with respect to the tape center are set one above and one below.

In FIG. 6, the intermediate value correctable burst error amount of Features 1 and 3.4 is indicated by diagonal lines. In the figure, the horizontal direction indicates the longitudinal direction of the tape, and the vertical direction indicates the scanning direction of the head.

The first characteristic, ■The fact that there is no set containing data of the same channel in a track, means that one head is caused by some reason as shown by the diagonal lines in Figure 6 (() (I?) (C). Even if playback becomes impossible (including temporary cases), if the other two heads operate effectively, there is an advantage that output can be made with complete intermediate value correction.

The second feature is that there are pairs of pairs for the center and edge portions, so all pairs have a uniform error rate compared to the error rate graph shown in Figure 2, which was explained in the conventional example. It has the advantage of being processed.

Therefore, the disadvantage that only a certain channel is always recorded/reproduced under conditions of poor error rate is eliminated.

The third and fourth features are that even if a burst error occurs in the longitudinal direction of the tape, the error data can be corrected to an intermediate value, as shown by diagonal lines in Figure 6 (D), (N), (F), and (G). It has the advantage of being able to output. However, in this case, as long as the fourth characteristic is satisfied, the third characteristic may not be satisfied.

As described above, in the scanning direction of the head, the area of each recording pattern is 173. In the tape running direction, even if 1/2 becomes impossible to detect data due to a burst error, data can be decoded using intermediate value correction.

Figure 5 shows the timing of the rad output and decoder for data scrambling reproduction set in the example of Figure 4.
The scale on the − layer in FIG. 5(A) indicates the rotation angle of the drum 9, and (■) indicates the processing timing of the decoder in accordance with this angle.

Figure 5 (a) shows the playback head output from heads A and B.

It is divided into C and shown in the range of approximately one and a half rotations 5406 of the drum 9. Further, (u) in the figure shows the processing timing of decoders b1...b6.

In the system of this embodiment, as is clear from FIG.
, B, and C is 120°.

The outputs of the three heads A, B, C are sent to six decoders bl...
...It will be assigned to b6. Decoder b1...
- The maximum processing time T0 required for signal processing of one PCM code area b6 corresponds to approximately 160° rotation time of the drum 9. Each decoder s, , b shown in FIG. 5 (17)
The processing sequence 2... allows each decoder b+, bz... to ensure uniform processing time from the setting position shown in FIG. 5(a). As shown in FIG. 5 (II), if we focus on one decoder, we see that it processes data from four channels as its input information channels, and the decoders that handle the aforementioned "pairs" are: , b2. ...
. . . similarly forms a “pair” and uses two decoders to generate 4
Channel data is now fully decodable. For example, the decoders S, bs; bz, b, 6; b+, and ba form pairs, respectively. As described above, according to the present scrambling method, it is easy to equalize the processing time in each decoder S, b2, etc., and the number of channels handled by one decoder can also be reduced. .

〔Effect of the invention〕

As explained above, according to the present invention, the non-uniformity of the error rate in the tape width direction, which is unique to the helical scan method, is made uniform for each data, and burst errors in the tape longitudinal direction and the head scan direction are corrected. is possible. Furthermore, with the scrambling method of the present invention, the processing time required to decode a set pattern is equalized, and the number of channels processed by one decoder can be set to a minimum.

[Brief explanation of the drawing]

Figure 1 is an explanatory diagram showing the formant of a conventional 3 mPcM 12-channel recording pattern, Figure 2 is a graph showing the error rate distribution of the helical scan method, and Figure 3 (
() (11) is an explanatory diagram showing the tape path and recording pattern diagram in one embodiment of the present invention, FIG. 4 is a detailed diagram showing data of the recording pattern of the embodiment, and FIG. 5 (()
(o) is a timing chart of data in each head and decoder of the embodiment, Figures 6 (a) to (g) are explanatory diagrams showing allowable burst errors of the embodiment, and Figure 7 is the distance of each area. Fig. 8 is an explanatory diagram showing m (tracks v
FIG. 9 is an explanatory diagram of the scrambled recording pattern of each data when m=5. 3...Trunk, 11. A, B, C...
・Het, 2... Recording pattern, 9...
・Drum, D... distance. Error rate list 3 recommendations (a) (b) joy q big

Claims (1)

[Claims] Having N (3 or more) information channels, M (≧2)
Rotating head system P with tracks and multiple heads
In the recording pattern of a CM multi-track system, the recording area of one rotation of the drum is divided into N areas.
A PCM code series, the information of each channel is divided into two pairs of EVEN data and ODD data, and each channel data set is combined with another channel data set, and the combination is divided into the above area. When distributing the data into m, it is prohibited to record the above-mentioned "pair" in one track, and when one track is divided into m, from the tape center to the tape width direction, Set the distance D for each area, and calculate the sum D_t of each distance of the recording area of the above "pair" as D_t={m/2 (m=even number), (m-1)/2or(
A data scrambling method in a PCM multi-track system, characterized in that it is set so that m+1)/2 (m=odd number).