JPS6016078A - Reproducing device of video signal - Google Patents

Abstract

PURPOSE:To prevent flicker and deterioration in picture quality by jumping the scanning position of a rotary head so as to conduct elimination and addition of a data in the unit of one frame in conducting compression and expansion of the reproducing time of a VTR. CONSTITUTION:The elimination and addition of a color video signal are conducted in the unit of one frame in compressing and expanding the reproducing time. Further, the timing for the elimination and addition is conducted by detecting that the reproduced phase of a reproduced video signal enters regions b2, b3, b6 and b5 advanced by a prescribed value or over in case of compression and by detecting that the reproduced phase enters regions a2, a3, a6 and a5 being delayed for a prescribed value or over in case of expansion. Thus, neither the fluctuation of a picture in the vertical direction nor the flicker exists because the elimination and addition of the video signal are conducted in the unit of frame.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a video signal reproducing apparatus such as a reproducing apparatus part of a VTR for broadcasting stations, and more particularly to a technique for changing and correcting the reproduction time in this reproducing apparatus.

BACKGROUND TECHNOLOGY AND PROBLEMS At broadcasting stations, programs, commercials, and the like are programmed and broadcast within a predetermined time frame. For this reason, it is often the case that the video is edited and recorded on a VTR and then played back and broadcast. By the way, in broadcasting sites, there are times when it is desired to broadcast urgent news or commercials that were not originally programmed within the time frame of the program.

In addition, there are cases where a commercial or the like that was originally programmed may be deleted due to an accident or the like. In response to these demands, broadcast sites are looking for a feature that can compress or expand the time axis of a video program, thereby making it possible to add or delete commercials immediately before broadcasting.

To achieve this, it is sufficient to be able to change and correct the playback time in the VTR. To this end, the tape speed at which the VTR plays back the tape may be changed according to the time axis compression rate or expansion rate. However, when changing the tape speed vT to a speed different from the normal playback speed in this way, if the rotational speed of the rotary head is constant, the scanning position of the rotary head with respect to the recording track gradually shifts. Furthermore, the scanning direction of the rotary head on the tape does not match the longitudinal direction of the recording track, which also results in tracking failure. However, if this tracking error occurs, it can be corrected using so-called dynamic tracking technology, which is used to playback at double or triple the normal playback speed, or to play back slow motion without noise on VTRs for broadcast stations. is possible.

In other words, SMP is an example of a VTR used at a broadcasting station.
TI! Type C Fu” - Record of Pine I. (1st is 1st to 3rd
If we consider a VTR that has the recording pattern 1j (() in the figure, the rotary head for video signals 1 (II, I) will record the video signal track The solid line (
2) Scan in a blur pattern. However, if the tape speed is faster than the normal speed, for example, 4 times the normal speed, the video signal rotary head moves the tape fl)-1- to the solid line (3) in Fig. 2.
), and when the tape speed becomes slower than the normal speed (for example, 10), the rotary head scans the tape (11-1-) as shown by the solid line (4) in Figure 2, Therefore, in this VTR, the rotary head for the video signal is moved in a direction perpendicular to the scanning direction by, for example, a piezoelectric element according to the voltage supplied to the piezoelectric element. Since scanning!IL traces can be predicted during commercial speed reproduction such as 2x speed or slow motion Pr generation such as 10x speed as described above, a tracking correction voltage is supplied to the piezoelectric element ζ,
The scanning trajectories shown by solid lines (3) and (4) are corrected to become scanning trajectories shown by dotted lines (5) and (6), respectively. This is dynamic tracking.

Therefore, even when compressing or expanding the playback time, the speed of the playback tape is controlled with high precision by a stable speed control, for example, by a microprocessor, and the scanning position of the video signal rotary head is Since the amount of deviation in the case where the image signal gradually deviates from the recording track can be predicted, the dynamic tracking control allows the video signal rotary head to correctly scan the recording track during reproduction.

In FIGS. 1 to 3, Tc is a control signal track, and Ts is a synchronization signal trunk recorded and reproduced by a rotary head different from the video signal head.

Incidentally, in this case, it is also necessary to reduce or add video signal information in accordance with compression or expansion of playback time. For example, to shorten the playback time by 1%, you need to delete 1 field out of every 100 fields, and to make it 1% longer, you need to add 1 field out of every 100 fields.For example, you need to make sure that the same field is not repeated twice. .

By the way, in the current dynamic tracking technology as described above, when the tape speed is different from the normal playback speed and the scanning direction of the rotary head no longer matches the longitudinal direction of the recording trunk Tv, the recording track Tv cannot be correctly tracked. Control is performed to just track the recording trunk where the amount of movement of the rotary head in the direction perpendicular to the scanning direction of the rotary head is the smallest with respect to the scanning locus of the rotary head. therefore,
This current dynamic tracking technology can be used as is to compress the playback time as described above.

When decompression is performed, deletion or addition of the video signal is performed by deleting or adding one field of the video signal by jumping the rotary head by one track.

In this way, the playback time is 1. l Two lines. Even after reducing or adding video signals during decompression, if the standard color television signal does not match the standard video signal (reference signal) at the broadcasting station,
The playback image becomes distorted on the screen.

However, when simply deleting or adding one field as described above, current television systems use an interlace scanning system, so the order of even and odd fields is determined by the signal of that one field. The reference signal is inverted from the position where it was deleted or added until the next deletion or addition is made.

That is, when compared with the reference signal of a broadcasting station, 100
/60#The field order is reversed every 1.67 seconds. If this field order is matched to the reference signal, the picture will move vertically on both sides by the interlace interval.

In high-speed playback of 2x speed or higher or slow playback of 10x speed or lower, one field is deleted or added for each field, and this is business luck (old sand cycle), so this field is deleted,
Although the image fluctuation due to the addition is not noticeable, the time axis compression and expansion of the playback time is around normal speed, that is, around 1x speed, for example, around 0.5 to 1.5x speed. Picture fluctuation due to deletion and addition jl1
1. This is noticeable as a flicker.

In order to prevent flicker caused by vertical shaking of this picture, the apparent Although it is possible to consider a method in which the horizontal line is positioned at the center between two adjacent lines, this would result in a vertical IW image resolution of ten, resulting in severe image quality deterioration.

Purpose of the Invention The present invention is directed to the above-mentioned points and provides a device that does not cause the drawbacks mentioned in item 1 when correcting changes in playback time in a video signal playback device. .

SUMMARY OF THE INVENTION In the present invention, when compressing and expanding the playback time, color video signal information is deleted or added frame by frame. Also deleted.

The timing of addition is determined by detecting that the playback phase of the playback video signal has advanced by more than a certain value when performing compression, and by detecting that the playback phase has delayed by more than a certain value when expanding. be. If you do the above, you can delete the video signal in frame units.

Since it is an additional image, there is no vertical image fluctuation and no flicker.

EXAMPLE Hereinafter, an example of the present invention will be described with reference to the drawings.

Note that the following example is a case where a device that plays back a recorded tape in SMPTE type C format is used as the playback device. That is, a single leaf yield video signal is recorded on the video signal recording track Tv, and the rotary head is rotated at a rate of one rotation for one field.

The bimorph 41 is arranged so that the video signal rotary head can move in a direction perpendicular to the scanning direction.
It is the same as described above that is attached to the horizontal moving means such as g, etc.

In this example, when the normal playback speed is 1, the playback time is supplemented by setting the tape speed to 0.75≦vT≦
Made between 1.25. That is, when the normal playback speed is 1x, the speed is between 0.15x and 1.25x.

In this case, the scanning direction of tape -1- of the rotating head does not differ much from that during normal playback, but
The scanning locus of the rotary head relative to the recording track gradually shifts. Then, if the scanning position of this rotating head is
Just tracking is performed by the head moving means in the same manner as in the case of dynamic tracking described above.

That is, for example, as shown in FIG.
When the speed is 1x or higher, the shifted scanning position is indicated by the solid line (7) in the same figure.
The rotary head is controlled so as to move from the position shown by the dotted line (8) to the position shown by the dotted line (8) in the figure. By the way, at playback time 0, the video signal rotating head extracts the signal from the tape fl.
) is determined in the width direction as shown in the figure. Therefore, when this head is moved in a direction perpendicular to the scanning direction in order to correct the scanning position of this head, since the recording track Tv is a diagonal track, just trasogging is performed at the corrected scanning position. Even if the signal is not extracted from all of the recording track Tv, as shown in FIG.
The length of is proportional to the amount of scanning misalignment of the rotary head.

This section d is a reproduced signal missing section, and the longer this section becomes, the more signals are missing in one field's worth of video signals. The above is a case where the tape speed is increased, but when the tape speed is decreased, the scanning position of the head gradually lags, so the vicinity of the end of the track in the scanning direction becomes a section where no reproduced signal is obtained from the rotary head. That is, in this case, in FIG. 3, the rotating head whose deviated scanning position was as shown by the solid line (7) is controlled so that it is brought to the scanning position where it is marked by the dotted line (9). Therefore, a section d' occurs in which no reproduced signal is extracted.

In the SMPTE type C format, the video track Tv includes the 14.5th horizontal line (hereinafter referred to as H) of the field to the 4th H (horizontal line is referred to as H) of the next field.
Second. In the 4th field, the 15th ■ (from 4.5) I
) is specified to record 252.5H signals. Therefore, when the tape speed of 7 tapes is increased to 1x speed or higher, the first section d of the track Tv is missing, so if this section dfJ < length < and becomes 3 H minutes or more, effective video data of the video signal will be missing. It turns out. Also, when the tape speed is reduced from 7 tapes to 1x speed or less, the section d' at the end of the track Tv is missing, so this section d'
If the length becomes longer than 4H minutes or more, valid video data will still be missing.

Looking at this in terms of the phase of the reproduced video signal with respect to the reference signal, when 0.75≦vT≦1.0, the reproduction frequency decreases and the internal vertical synchronization signal phase (Track 1) The vertical synchronization signal is recorded on the track Tv. (equivalent to the playback phase if the
When the value is 1.25, the reproduced vertical synchronizing signal phase gradually advances.

By the way, in the SMPTE type C formant, the intervals d and d for a deviation of one trunk pitch τp are
For example, the length of ' is 2.5H in the case of an NTSC signal. Therefore, the reproduced vertical synchronization signal phase lags the reference signal by more than 1.25τp (=3H),
Moreover, if the process proceeds for more than 1.5τp (#4)(), valid video data will be missing.

Based on the above considerations, and with some safety in mind, the allowable range of delay for the reproduced vertical synchronization signal phase with respect to the reference is -1.
, 25τp”1.0τp.

In this example, when the reproduction vertical synchronization signal phase (equal to the reproduction phase) exceeds this tolerance range, the scanning position of the rotary head is changed by one frame of the video signal to SMPT[! In Type C format, the video signal information is deleted by jumping two tracks Tv.

make an addition.

3 2 As an example, it is carried out as follows.

In other words, when the tape speed v is set to 1 during normal playback, and when 0.75≦V≦1.0, the maximum delay amount d1 per field (one track) is
is 0.25r p when V T = 0.75c7)
There is. On the other hand, since the maximum allowable delay is 1.0τp, the detection phase for making the rotating head track jump is the sum d2(at ≦d2) is set.

In this example, 0.7τp (1..
(equivalent to a delay of 75H). Tape speed 7 on horizontal axis
In FIG. 4, where the relative delay phase (unit pitch τp) is plotted on the vertical axis, the detected phase is a4as, and the phase region ax a2as a4 delayed from this is the detected phase region. When the reproduction vertical synchronization signal phase enters this area, the rotary head moves one step in the direction in which the rotating head advances relative to the tape running direction in synchronization with the reproduction vertical synchronization signal phase.
It is caused to jump by one frame (2τp). Then, the reproduced vertical synchronizing signal in the area al a2 a6 4 a4 starts scanning the trunk of the next field from the corresponding phase in this area b1 b2 b5 b4, as if it were in the area b1 b2 b5 b4 advanced by 2τp. Then, from this point on, the delay starts to gradually start again, and when the reproduction vertical synchronization phase enters the area a1a2as a4, the track is jumped again. This will be repeated below.

In this case, when the rotating head is caused to jump the scanning position, the jump is one frame, so no reversal of odd and even fields occurs, so the field phase of the VTR playback signal matches the field phase of the reference signal. If they match, no problem will occur if the signal is black and white.

Next, the tape speed vT is 1. Q≦V T ;ii,
Consider the case of 1.25.

In this case, the reproduced vertical synchronization signal phase will gradually advance, but the maximum advance amount d3 per field is V, -
In the case of 1,25, it becomes 0.25τp. Therefore, we now set the maximum allowable advance phase to 1.3τp (section rr in Figure 4).
, phase amount d4 (d4=d2≧
The phase delayed by d3) is set as the detection phase. In this example, it is set to 1.0τp (corresponding to a delay of 2.5H in the case of an NTSC signal). This becomes b2b3 in FIG. Therefore, the phase area b2b3bebs becomes the detection phase area, and when the reproduced vertical synchronizing signal phase enters this area b2b3 b6bs, the rotary head moves 1 in the direction behind the tape running direction in synchronization with the reproduced vertical synchronizing signal phase. The signal is jumped by one frame, that is, 2τp. Then, in the next field, it is as if the previous reproduced vertical synchronizing signal phase is delayed by 2τp in the area a2 a3as
This area 82aa aca as if it were in a5
Scanning of the track starts from the corresponding phase in s, and from this point on, it starts to advance gradually again, and from then on, the same operation as above is repeated.

Even in this case, since the rotating head jumps in units of frames, odd and even field inversion does not occur.

5. In actual broadcasting, it is necessary to match the phase of the reproduced signal, which is leading or lagging, with the reference signal to make a signal that conforms to the broadcasting standards, and this is done by a TBC (time base collector).

In order to perform a track jump for one frame of the rotary head, it is necessary that the odd and even field relationships of the reference signal and the 111 raw video signal match.

Therefore, if the relationship between the odd and even numbers of fields does not match, if vT≦1.0, the reproduced vertical synchronizing signal phase will be delayed by one field (one track pitch τp) when entering the area al a2 asa4. If the rotary head jumps so as to advance by
When entering 3b6bs, the rotary head is jumped so as to be delayed by one field (one track pitch τp), thereby matching the field of the reproduced signal and the field of the reference signal, and then one frame of 7 6 as described above. This is to make a unit track jump.

As described above, it is possible to compress or expand the playback time, and in this case, it is possible to prevent the deterioration of meat quality due to odd and even field inversion. Moreover,
In this case, the amount of video information deleted or added is
The time is approximately 1/30 second for the C format and approximately 1/25 second for the PAL format, but considering that the editing power of movie film is 1/24 second per month, no problems such as flickering occur.

By the way, the above is a case where only the luminance signal is considered. Therefore °ζ, recording of black and white video signals.

The reproduced and color video signals are recorded as components (separated into a luminance signal and red and blue color difference signals, for example), and these are recorded as n, ? -1: In the above case, the change in regeneration time can be properly corrected by the above method, but if the color video signal is recorded as a composite signal (composite color video signal) and you want to play it back, the following Further processing will be required.

For example, to explain the case of an NTSC composite signal, first, the color sub-Ill! The phase of the transmitted wave becomes a problem.

That is, horizontal frequency fH1 color subcarrier frequency fsc.

As is well known, the relationship between the field frequencies fv is as follows: (1) The color subcarrier per field enters every 5+718.75 cycles, delays by +cycles per field, and returns to the original state in 2 frames.

Therefore, when one frame of video signal is deleted or added, from that point on, the phase of the color subcarrier becomes inverted with respect to the reference signal. That is, the fifth
Figure A shows the field number when considering repetition of the color subcarrier phase of the reference color video signal, and Figure B shows the phase of the color subcarrier at the beginning position of the first horizontal line of each field. , in C of the same figure, the tape speed VT≦
The reproduced signal phase and its field number at the time of 1 are shown in the same figure, and the corrected reproduced field number after correction (after TBC is also done) is shown in the same figure. The phase of the color subcarrier at the position shown in FIG.

As is clear from FIG. 5A to 1Σ, the phase of the reproduced color sub-primary transmission wave is reversed by a track jump of one frame until the next track jump of one frame is made. I understand.

In addition, Fig. 5F shows the external signal phase and its field number when the tape speed vT≧1, and Fig. 5G shows the complement i at that time.
The supplementary reproduction field number of Ef& is shown in FIG.

It is clear that in this case as well, the phase of the color subcarrier is inverted for every jump of one frame.

This inversion of the phase of the color subcarrier means that when trying to obtain a signal that conforms to the broadcasting standards by connecting a TBC to the latter stage as described above, the horizontal position of the screen will shift. It becomes a deficiency.

In other words, the horizontal position is determined by the phase of the water-IP synchronization signal, but since the horizontal synchronization signal is usually undesirable due to jitter, about 12
Stable true Tll of burst signal inserted by cycle
is determined in TBC by the phase of which cycle. However, since the phase of this burst signal is reversed by one frame jump, the +cycle demarcation position of the burst signal is used as a reference for the horizontal position, and the screen shifts in the horizontal direction.

In order to prevent this, after the burst signal serving as the horizontal phase reference is frame-jumped, it may be further inverted and used during the field period in which it is inverted with respect to the reference.

The above prevents the horizontal position of the screen from shifting, but
If you only invert the burst signal phase, the color (Bi) phase will still be inverted. Therefore, once the reproduced composite signal is separated into a luminance signal and a chrominance signal, and the chrominance signal ()γ phase is inverted, the luminance signal and then add it again to return it to a composite signal.

Several examples of the signal processing system of the present invention apparatus taking the above-1- into consideration are shown below.

1 0 Figure 6 is an example, 00) is S M 11 T Ii
This figure shows a playback unit of a VTR compatible with Type C format.

In V T RG11l, (11) is a rotary head for video signals, which is attached to a head scanning position moving means (12) made of a piezoelectric element such as a bimorph, so that the rotary head (11) is phantom in the scanning direction. It is movable in orthogonal directions, and is controlled by a control signal from a tracking control circuit (13) so that reproduction is always performed in a just tracking state.

The playback output from the rotating head (11) is sent to the playback circuit (14).
The signal is FM demodulated at , and a reproduced color video signal is obtained from this.

Further, the repeating video signal is supplied to a synchronizing signal forming circuit (15), and a signal VD equal to the reproduced vertical synchronizing signal is formed from, for example, an equalization pulse in the reproduced signal from the track Tv. This signal vD is fed (jt) to the phase comparison circuit (16), and the vertical synchronization signal VR in the reference signal is also supplied to this phase comparison circuit (16), the phases of both 9t are compared, and the ratio 2 is the aperture output. is supplied to the jump detection circuit (17).

This detection circuit (17) includes a reference phase level setting circuit (1
The reference level from 8) is supplied. This setting circuit (1
8) is the above-mentioned area 81a2 according to the tape speed vT.
as a+ and its velocity v in the area b2b3be bs
The output at a level corresponding to the lowest detected phase at T is supplied to the detection circuit (17) as a reference level.

Therefore, when the phase of the reproduced vertical synchronization signal Vo deviates from the phase of the reference signal vR by a predetermined amount, the detection circuit (17) obtains a signal J that rises to a high level, for example. This signal J is supplied to a tracking control circuit (13) to cause the rotary head (11) to jump forward or backward by one frame with respect to the tape running direction.

The output from the reproduction circuit (14), ie, the reproduced color video signal from VTRQl, is supplied to the TBC (20).

That is, the reproduced color video signal is converted to an A/D converter (2
1) and a timing signal generation circuit (2).
2). Then, the reproduced color video signal signal is converted into a digital signal in the A/D converter (21) by the write clock signal CKw which is made to have the same jitter as the reproduced color video signal from the timing signal generation circuit (22), Same clock signal CKw
is written into memory (23) by . Then, a digital signal is read out from this memory (23) using a constant frequency read clock signal CKR without jitter from a timing signal generation circuit (22), and jitter is removed.

Further, the jump detection signal J from the VTROL is supplied to an inverted flip-flop circuit (30), and the output of this flip-flop circuit (30) changes state every time the rotating head (II) jumps one frame. Invert.

This output from "C1" is output from the timing signal generation circuit (22
), and during the period when the phase of the color subcarrier is inverted, the burst signal in the reproduced color video signal is inverted, and the horizontal position of both sides is set by this inverted burst signal. .

3. The jitter-removed digital color video signal from the memory (23) is supplied to a Y/C separation circuit (24) and separated into a luminance signal Y and a color signal C. The luminance signal Y is supplied as is to the synthesis circuit (25), and the color signal C
The one inverted by the phase inversion circuit (26) and the one with the same phase are selected in the switch circuit (27) by the signal from the flip-flop circuit (30) and supplied to the synthesis circuit (25). be done. That is, during the period when the color subcarrier is inverted due to one frame jump of the rotating head, the switch circuit (27)
The inverted color signal from the phase inversion circuit (26) is taken out and combined with the luminance signal Y. Then, the output of the synthesis circuit (25) is returned to an analog signal in the D/A converter (28), and is led out to the output terminal (29).

In addition, in the TBC (20), as mentioned above, there are sections d and d' in which the playback signal cannot be extracted by the rotary head.
This also compensates for the horizontal sections that have dropped out.

5 4 In this example, the color signal and luminance signal are separated by A/
It may also be done between the D converter (21) and the memory (23).

FIG. 7 shows an example of such a case, in which the output signal of the A/D converter (21) is supplied to the Y/C separation circuit (24') and is separated into a luminance signal Y and a color signal C, and the luminance signal Y is The color signal C is stored in the luminance signal memory (23Y) and the phase inversion circuit (26Y).
) and the switch circuit (27'), the phase of the color subcarrier is made appropriate according to the output of the flip-flop circuit (30), and then each is written into the color signal memory (23C). In this case, the timing signal generation circuit (
Writing and reading are performed using the write clock signal C~ and read clock signal CKH from 22), and jitter and ivy correction is performed, and dropout correction is performed in these memories (23Y) and (23C) as described above. The same is true, and the horizontal positioning of the screen is also done in the same manner as in the example of FIG.

The luminance signal Y and color signal C read out from these memories (23Y) and (23C) are synthesized by a synthesis circuit 6 (25), and converted into an analog color video signal by a D/A converter (28). It is something that

In the second example below, the horizontal positioning of the screen may be performed using a reproduced horizontal synchronization signal, but even in that case, color signal inversion processing is necessary.

FIG. 8 shows still another example of the present invention. In this example, a reproduced color video signal from a VTR (101) is directly converted into a luminance signal and a color signal in a Y/C separation circuit II (24"). and the color signal is switched between inversion and non-inversion in the same manner as described above using the outputs of the phase inversion circuit (<26"), the switch circuit (27"), and the flip-flop circuit (30). , then the synthesis circuit (3
In 1), it is combined with the luminance signal to obtain a composite signal that is a continuous color subcarrier, and this is supplied to the TBC (20') to perform jitter correction and dropout compensation.

In the above example, if a VTR with large jitter is used, another memory for jitter correction is installed between the synthesis circuit (25) and the D/A converter (28) or the Y/C separation circuit. It may also be provided between the A/D converter and the A/D converter.

Although the above example has been explained using the case of playing back a recording tape in the SMIITE type C format, the present invention is not limited to this, and also applies not only to MTSC format video signals but also to PAL and 51 ICAM format video signals. Of course, it can also be applied.

Effects of the Invention As described above, in this invention, when compressing or expanding the playback time in a VTR, the scanning position of the rotary head is adjusted so that data of the video signal is deleted or added in units of one frame. Since the image data is jumped, there is no flicker caused by image fluctuation in the vertical direction or deterioration of image quality, which occurs when video signal data is deleted or added in units of one field in the conventional method.

Therefore, the feature desired by broadcasting stations, ie compression and expansion of playback time, can be very well realized.

[Brief explanation of the drawing]

7 Figures 1 to 3 are diagrams for explaining the scanning locus of the rotary head on a recording tape of SMPTB type C format, Figure 4 is a diagram for explaining an embodiment of the present invention, and Figure 5 is A diagram for explaining the relationship between field order and color subcarrier phase when this invention is applied to reproduction of a color video signal, FIG. 6 is a system diagram of an example of this invention, and FIG. FIG. 8 is a system diagram of still another example of the present invention. α [is a VTR, (11) is a rotary head for video signals, (1
2) is a head moving means, (13) is a tracking control circuit, and (17) is a jump position detection circuit. 9 8 JP-A-Sho GO-16078 (11)

Claims (1)

[Claims] When the tape feed speed vT during normal playback is set to 1, the device changes the tape feed speed between 0.5<V<1.5 to achieve +IG output/open compression/expansion, A head moving means is provided that can move the rotary head rotating at a constant speed in a direction perpendicular to the scanning direction, and the head moving means allows the rotary head to move even when the tape speed differs from the normal speed. In addition to always scanning the recording I/rack correctly, when the speed vT is V7<1, it is detected that the playback phase is delayed by a certain value or more from normal playback. The means causes the rotary head to jump in the direction in which it advances with respect to the tape so as to scan a position where one frame of the video signal is skipped, and when V7>l, the reproduction phase is kept at a constant value than during normal reproduction. When it is detected that the video signal has advanced two times, the moving means causes the rotary head to jump so as to scan a position where one frame of the video signal is added in a direction behind the tape. Video signal playback device.