Note: Descriptions are shown in the official language in which they were submitted.
i ~
SO 1361
BACKGROUND OF THE INVENTION
This invention rela-tes qenerally to a method and
apparatus for recording a digi-tized video siqnal on a ma~netic
-tape and, more Particularly, is directed -to a me-thod and
apparatus for recording a digitized video signal on a
magnetie tape with a high recording density.
Conventionally, apparatus for recordin~ a
video slgnal on a magne-tic ta~e have been of the analo~,
rather than digltal, -type. However, there has been a
reeent turn towards development of dlgltal video -tape
reeordexs (VTR). Di~ital VTRs have a very hiqh picture
quality, whieh enables multlple generatlon dubbing with
vlrtuallv no pleture im~airment. .Further, dlgital VTRs provide
adjustment free eircuits and self-diagnostic systems which
enable easier mai.ntenanee and higher reliability.
With digital VTRs, an analoq video signal is
eonverted into digital form by an AJD tanalog-to-digital)
eonverter. In partieular, the analog video signal is sampled by
eloek pulses having a sampling frecruency which may be, for
e~ample, ~ESc, where fse is the color sub-earrier fre~uency
of the eolor video signal, resulting in -the analog vicleo
signal bein~.eonverted lnto a digitized video signal. eomprised
of 8-bit words~ The digitized signal is also coded bv
an error eontrol encoder so that errors may be corrected
and eoneealed on playback and, it is further coded by a
ehannel encoder to achieve high den.sity digital recordinq.
-2-
The eoded digitized signal is then recorcled on a magnetic
tape by means of a recording amr?lifier. 130wever, it should
be appreciated from -the above that the recordinq bit rate,
that is, the rate of oceurrence of eaeh bit of the cdigltized
video signal, is extremely high. For example, in the above-
deseribed embodiment, where -the color sub-carrier frequency
fse = 3.58 MHz, the reeording bit rate is equal -to ~fsc times
the number of bits per word. In other words, the reeording
bit rate is ohtained as follows:
Bit rate = ~ x 3 58 x 106 x 8 = ll~.6 Mb/s.
Because of such high recordinq hit rate, the digit~zecl video
signal is not suitable for recording in a sin~le recording
ehannel.
Accordingly, it has been proposed to separate the
digitized video signal into at least two separate channels
prior to recording it on a magnetie tape so as to reduce the
recGrding bit rate per ehannel. Typieally, a maqnetic head
is assoeiated with eaeh ehannel and all oE the mac~netle heads
are ali~ned to reeord the respeetive channels on a magne-tic
tape in parallel -tracks extending obliquely on the tape. In
order to separate the digitized videc signal into, for example,
two ehannels, an interface is provided which distributes
alternate 8-bit words of the digitized video signal into the
respective channels.
In recording the cligitized vicleo signal in the
parallel tracks, it is desirable to increase the siqnal-to-noise
(S/N) ratio so that, during reproduetion, a video ~icture oF
--3--
~2~
high quali-ty can be obtained, while at the same time,
reducing the amoun-t of tape consumption by recordin~ the
digitized signal with a high density. Tt should be
appreciated that these two ob~ections are contrary to one
another. For example, as the track width is decreased so
as to obtain such hic~h density recordi.ng, the S/N ratio of
the video signal reproduced from the tracks deteriorates.
In like manner, as -the track width is increased which resul-ts
.in higher tape consumption, -the S/N ra-tio lncreases.
Therefore, in such previously proposed a~paratus,
guard bands have been provlded between adjacent ones o:E the
parallel tracks recorded on the magnetic tape so as to avoid
cross-talk interference between such adiacent tracks, resulting
in a higher S/N ratio. ~lowever, when the track width is ~0~ m,
for example, the wldth o the ~uard band between each of the
adjacent tracks must be at least ?0~ m, resul-ting in a high
tape consumption. IE, on the contrary, the track wid-th is
made narrower, tracking errors are apt to occur during the
reproduction operation, wherein the heads do not accurately
trace the recorded tracks, resulting in a deterioration of
the S/N ratio. Also, with a reduction of track width, there
necessarily is a reduction in the width of the guard bands,
resulting in increased cross-talk interference (noise) from
adjacent tracks.
OBJECTS AND SUM~ARY OF THE INVENTION
Accordingly, it is an ob~ect of this invention to
provide a method and apparatus for recording a digitized
video signal on a ma~netic tane that avoids the above-described
~8;~
dif:Eiculties encountered with -the prior ar-t.
It is another object of this invention -to provide a
method and apParatus :Eor recordinq a diqi-tized video siqnal
in which the digitized video signal is sequentially
distributed to a plurality of channels and then the
digitized video signal in each channel is recorded in
a plurality oE parallel tracks ex-tending obliquely on a
magnetic tape without guard bands between adjacent tracks.
It is still another object oE this inven-tion to
provide a method and apparatus for recordinq a digitized
video signal in which the video signal, upon being reproduced,
has a high signal-to-noise (S/N) ratio.
It is yet another object of -this invention to
provide a method and apparatus for recording a digitized
video signal in which the digitized video signa:L is recorded
with a high recording density in a plurality of parallel tracks
exten`ding obliquely on a magnetic tape so as to reduce tape
consumption.
I-t is a further object of this inven-tion to provide
a method and appara-tus for recoxdi.ng a digitized video signal
.in which the digitized video signal is recorded in a plurali-ty
of parallel tracks extending oblicluely on a magnetic tape and
having a guard band between adjacen-t tracks, with -the digitized
video signal in at least some of the tracks being recorded
with an azimuth angle which is di~Eerent from other ones oE
the tracks.
In accordance with an aspect of thi~ invention,
apparatus for recording a video si(~nal on a magne-tic tape
~2
includes means for converting the video signal into digi-tal
form; means for distributing respective portions o:E the
digi.tized video signal to at least -two channels; and means
for recording the respective portions of the digitized video
signal in a plurality of parallel tracks extending obliquely
on the magnetic tape without guard bands between a-t least
some adjacent ones oE the parallel trac]cs and with the portions
of the digitized video signal in some of the parallel -tracks
being recorded with an azimuth angle which is di:Eferent from
the azimuth angle in other ones of the parallel tracks.
In accordance with another aspec-t of this invention,
a method of recording a video signal on a magnetic tape
includes the steps of converting -the video signal into
digital .~orrn; distributing respective portions of the digitized
vldeo signal to at least two channels; and recording the
respective portions of the digitized video signal in a plurality
of parallel tracks extending obliquely on the magne-tic tape
without guard bands between at least some adjacent ones o:E
the parallel tracks and with the portions of the digi-tized
video signal in some of the parallel trac]cs being recorded
with an azimuth angle which is diEferent :Erom the azimuth
angle in other ones of the parallel trac]cs.
In accordance with a further aspect of this
invention, the diqital video s.ignal, prior to recording, is
code converted, for example, by an 8-to-10 code conversion
system, in order to reduce low frequency spectrum components
of the digitized video signal so as to improve the S/N ratio
of the video signal when reoroduced.
--6--
~Z2~
More particularly there is provided:-
Apparatu~ for recording a video ~ignal on amagnetic tape comprising:
means for converting the video signal into digital
form;
-means for distributing respective portions o the
digitized video signal to at least two channels, each o~ said
at least two channels including code converting means for code
converting the respective portions of said digitized video
signal distributed to the respective channel so as to reduce
low frequency components of said digitized video signal; and
means for recording the respective portions of said
digitized video signal in a plurality of parallel tracks
extendiny obliquely on said magnetic tape without guard band~
between at least some adjacent ones of the parallel tracks and
w.ith said portions of the digitized video signal in adjacent
ones of the parallel tracks without guard bands therebetween
being recorded with different azimuth angles.
~ There is also provided:-
A method of recording a video signal on amagnetic tape comprising the steps of:
converting the video signal into digital form;
distributing respective portions of the digitized
video signal to at least two channels;
code converting the respective portions of said
digitized video signal distributed to each channel so as to
reduce low frequency components of said digitized video signal;
and
recording the respective portions of said digitized
video signal in a plurality of parallel tracks extending
obliquely on said magnetic tape without guard bands between
at least some adjacen~ ones of the parallel tracks and with
said portions of the digitized video signal in adjacent ones
of the parallel tracks without guard bands therebetween being
recorded with different azimuth angles.
-6a
22~
There is further provided: '
A method of recording a video siynal on a mag-
netic ta~e comprising the steps of:
converting the video signal into digital form;
distributing respective portions of the digitized video
signal to at least three channels, three sequential ones of said
parallel tracks including a field interval of video information;
and
recording the respective portions of said digitized
video signal in a plurality of parallel tracks extending obliquely
on said magnetic tape with a guard band provided after every
three sequential ones of said parallel tracks and without any
guard bands between adjacent ones of the remaining tracks, and
with the digitized video signal in the center track of every
three sequential ones of said parallel track~ corresponding to
a field interval and being recorded with an azimuth angle which
is different from the azimuth ~ngle in the remaining ones of the
parallel txacks.~
-6b-
The above, and other, objects, features and
advanta~es of the present i.nvention, will be apparen-t from
the Eollowlng detailed description which is to be read in
connection with the accompanying drawings.
~RIEF DESCRIPTION OF THE DRAWINGS
Fig. 1 is a schematic diagram used for explaining
the recording of a digitized video signal in parallel -tracks
with guard bands between adjacent tracks, accorcling to a
prev:iously proposed method of recording a digi-tized video
signal;
Fig. 2 is a graphical diagram illustrating the
cross-talk characteristics of the previously proposed method
of Fig. 1 and the method according to this invention;
Figs. 3A and 3B are schematic diagrams illustrating
the tracking hy A magnetic head with apparatus accordinq -to
this invention and with previously ~roposed apparatus for
recording a digitized video signal;
Figs. ~A-4D are waveform diagrams of various digital.
code converting formats;
Fig. 5 is a graphical diagram of the Erequency
spectrum density for the various formats shown in Figs. ~A-4D;
Fig. 6 is a graphical diagram oE the frequency
spectrum density illustrating the reduction of low .frequency
components by means of an 8-to-1~ code conversion system;
Fig. 7 is a block diagram illustrating a recording
section of a digital video tape recorder (VTR) embodying this
invention;
Fig. 8 is a hlock diagram illustrating a reproducing
2@~D
section of a digital video tape recorder (VTR) which is
complementary to the recording section of Fig. 7,
Fig. 9 is a schematic diagram illustrating the
positional and azimuth angle relationship be-tween the four
magnetic heads of the recording and reproducing sections of
Figs. 7 and 8;
Fig. 10 is a schematic diagram of a rotary head
assembly included in the digital VTR of Figs. 7 and 8;
Fig. 11 is a schematic plan view of a section of
magnetic tape showing tracks in which the signals are recorded
by the recording section of Fig. 7 and showing the rela-tion-
Silip between the azimuth angles in adjacen-t tracks;
Figs. 12, 13 and 14 are schematic diagrams to which
refererce will be made in explaining the digitiza-tion and
code arrangement of a video signal for use in a digi-tal VTR
embodying this invention; and
Fig. 15 is a schematic plan view of a section of
magnetlc tape illustrating the relationship between adjacent
tracks in accordance with another embodiment of -this invention.
DET~ILED DESCRIPTION OF THE' PREFERRED EMBODI~lENTS
In order -to facilitate a bet-ter understanding of
the presen-t inven-tion, there will first be described the
condi.tlons for di.gital recording of a color video signal
with a high signal-to-noise (S/N) ratio and a high recording
density.
Where a digitized video signal is transmitted, a
tolerable bit error rate in the transmission of the digitized
--8--
2~9
signal is 1 x 10 7. Since the S/N ra-tio oE a transmission
path (where the signal is measured by i-ts pealc-to-peak value
and the noise is measured by an effective value) is more -than
20 dB when the bit error rate is sligh-tly less than 1 ~ 10 7,
the S/N ratio of a digitized color video signal obtained,
during reproduction from a digital VTR, must be larger than
20 dB.
In addition to the requirement for a suEEiciently
hicJh S/N ratio, it is also desirable to reduce the tape
consumption so as to obtain maximum utilization oE the
magnetic tape. This, of course, means that the digital
video signal must be recorded with a high bit denslty. In
order to obtain such high bit density recording, the number
of recorded bits per unit area of.tape S (recording bit density)
mùst be high, in which the recording bit densi-ty S is
expressed by the following equation:
S = L T ....(1)
where L is the line bit density, that is, the number of recordecl
bits per unit length in the leng-tl-lwise direction of the
track, and T is the track density, that is, the number of
recorded bits per uni-t length in the wi.dthwise direction of
the track. Generally, as the value of -the line bit densi-ty L
increases, a short wavelength for recording must be utilized.
Assuming that the magnetic layer on the tape is
sufficiently thick, it has been determined that the number
of magnetic particles which are activated so as to change the
2~
magnetic flux supplied -to the reproducing head increases
approximately in propor-tion to the square of the recording
wavelength utili~ed. Further, the signal vol-tage generated
at the reproducing head increases in proportion to the number
of activated magnetic particles, while the noise voltage
generated at the reproducing head increases in proportion
to the square root of the number of activated magnetic
particles. In other words, the signal voltage generated at
the reproduced head increases In a proportional manner -to the
square of the wavelength and the noise voltage generated at
the reproducing head increases in a proportional manner -to
the wavelenyth. Thus, if it assumed that the source of
noise results only from the tape, tha-t is, the activated
magnetic particles thereon, the S/N ratio of the reproduced
digitized signal increases in a proportional manner -to the
w~avelength. Further, -the S/N ratio for the amplifier system
oi the VTR is also proportional to the waveleng-th. I-t
should therefore be appreciated that if the track width and
the relative speed between the reproducing head and the tape
are constant, the S/N ratio increases as -the recordlng wave-
length increases. However, it should be appreciated -that this
is contrary to the condition of high bit density recording
where it is desirable to utilize a short wavelength in order
to increase the line bit density L,and consequently, to -thereby
increase the recording bit density S.
In regard to the track density T, -the signal voltage
and tape noise voltage generated at the reproducing head
each decrease in a proportional manner to reductions in the
--10--
~2~
track width W. However, if the noise is generated only from
the tape, the noise voltage generated at the reproducing head
is only in proportion to the square root of the track width W.
In such case, the S/N ratio of -the reproduced digitiz.ed siynal
is proportional to the square root of the -track wid-th W.
In regard to the noise from the VTR, the induc-tance
of the reproducing head is approximately proportional to the
width oE the reproducing head, that is, to the track wid-th W.
When the inductance of the reproducing head is constant, the
num~)er of turns of windings on the head are inversely
proportional to the square root of trac]c width W. Further,
the ma~netic flux linked with the windings o:E the head is
proportional to the track width W. It should therefore be
appreciated tha-t, with the inductance of the reproducing head
maintained at a constant value, the voltage induced in -the
reproducing head is proportional to the number of turns N times
the magnetic flux ~b intersecting the windings. In other
words, the voltage E induced in the reproducing head is
proportional to the square root of the -track wid-th W. Further,
if the inductance of the reproducing head is cons-tan-t, -tlle
n~ise generated by the reproducing head amplifier is also
constant. Thus, assuming tha-t the source of noise only resul-ts
from the reproducin~ head amplifier, the S/N ratio of the
reproduced digitized signal is proportional to the square root
of the track width W. If the generated noise from the tape
and from the reproducing head amplifier are independent of each
other, the S/N ratio of the reproduced digitized signal, as
a result of the combined noise from the tape and the
~8Z~Q~
reproducing head ampliEier, is proportional to -the square
roo-t of -the traek width W. In other words, reduction of the
traek width ~ so as to increase the -track density T results
in a deterioration of the S/N ra-tio.
It should be appreciated from the above that
the recording bit density S is increased by redueing the
traek width W so as to increase the track densi-ty T and by
utllizing a short recording wavelength so as -to increase the
line bit dènsity L. However, such conditions result in a
deterloration of the S/N ratio. I-t should therefore be
appreeiated that the eonditlons for increasing the S/N ra-tio
while inereasing the reeording bit density S are eontrary -to
one another.
In order to eompensate for the above, previously
proposed digital video tape recorders have recorded the video
signal in a plurality of parallel tracks extending obliquely
on a magnetie tape with guard bands between adjacent tracks, as
shown in Fig. 1. This results in a reduction of cross--talk
noise interference caused by leakage magnetic flux from
adjaeent traeks with a eonsequent increase in the S/N ratio.
OE eourse, SUCIl eross-talk interferenee relates only to the
noise generated by the tape and no-t crom the reproclucing head
ampl.ifier. However, in such ease, if the traclc density T is
inereased, that is, the traek width W is decreased, so as to
increase the reeording bit density S, the guard bands between
adjaeent traeks are also reduced. This results in increased
eross-talk noise interference from adjacent tracks. Also, in
sueh ease, when the track width W beeomes too narrow, tracing
-12-
z~
of the tracks by the reproducing heads becomes difficul-t so
that lncorrect tracing ls apt to occur with a consequent
deterioration of the S/N ratio.
Before proceediny further, the above-mentioned
cross-talk interference from adjacent tracks will be
discussed. Referring firs-t to Fiy. 1, there is shown a
reproducing head 1 having a width W and a plurali-ty of
parallel recorded tracks 2 with guard bands between adjacent
tracks. The width of the tracks is equal to the head width W
and a magnetized width ~W caused by fringe flux and which
is equally divided on both sides of each track so that a
fringe flux width ~ W/2 extends in the widthwlse direction on
both sides of each track. Given that the width of each ~uard
band is equal to x, the wavelength of the recorded signal is
~ , the level of the desired or true recorded signal is E, and
the level of the cross-talk signal is ~c' the cross-talk
interference component Ct can be expressed by the following
equation:
Ct = 20 1Og(Ec/E) = A -~ B ~x/~ dB ---(2)
where
b aw -bW + ~W b x
A = 20 log[ K . ~b e (1-e )e ] ...... (3)
K - W + 2~ (l-e b -~) ... (4).
Further, it is assumed that x ~> ~ W and, by experiment, it
-13-
has been determined tha-t ~W - 0.67 ~ , b ~ 6.~ and
B - -60.
If values for the track width W and the guard band
width x are chosen as 40~m and 20~ m, respec-tively, from
equation (2), when the relative speed of the reproducing head
to the tape is 25.59 m/sec, the Erequency characteristic for
the theoretical cross-talk component is shown by curve Cl in
E'iy. 2. In other words, curve Cl represents the cross-tal]c
interference resulting from the two end tracks in Fig. 1
when head 1 traces the center track of Fig. 1. I-t should be
appreciated from this curve that the level of the cross-talk
interference is subs-tantially increased for low fre~uency
components of the video signal.
It should further be appreciated that the
reduction of guard band width x and, of trac]c width W,
neeessarily results in -the tracks being closer to one ano-ther
so as to result in increased cross-talk interference. Therefore,
there is a limit to the reduction of guard band width x in
order to provide that a picture can be reproduced with good
~uality.
Further, as previously discussed, as the trac]c width
W is decreased, it becomes difEicuLt to accura-tely trace
each track. In o-ther words, reproducing head 1 is apt to
deviate from the desired path of the recording -track. This,
of course, results in a subs~antial increase in cross-talk
interference from adjacent tracks. Although the tracking
aecuracy can be improved by various servo techniques, it is
fundamentally determined by the mechanical accuracy of the
system which cannot be accurately controlled. In this manner,
--1'1-- `
~8;2~
it should be appreciated that because of mistracking and
cross-talk interference from adjacent tracks, -the track width
W and guard band width x can only be reduced by a cer-tain
amount in order to increase the recording bit density S.
Referring now to Figs. 2 and 3, the basis of
the present invention will now be described, which is
aimed at obtaining a high recording density while, at the
same time, substantially increasing the S/N ra-tio. In
accordance with -the present invention, the digi-tized color
video signal is distributed to a plurality of channels and
the slgnals from -the channels are recorded by recordiny heads
associated wi-th each of the respective channels in adjacen-t
parallel tracks extending obliquely on a magnetic -tape with
the longitudinal edges of the adjacent tracks being in contact
with each other so as to eliminate any guard bands -therebe-tween.
Further, the digitized signal is recorded so tha-t the azimu-th
angles in adjacent tracks are differen-t from one anotherl with
the azimuth angle for each track being defined by the angle
hetween the direction of the air gap of the recording head
utilized and a reference direction, for example, the directior
perpendicular to the longitudinal direction of -the track.
Preferably, the azimuth angles ~ in adjacent tracks are equal,
hut opposite, as shown in Fig. 3A. During reproduc-tion,
reproducing heads with air gaps having the same direction as
those of the respective recording heads trace the respective
tracks to reproduce the digitized video signal therefrom. By
utilizing such arranyement, cross-talk interference between
adjacent tracks is substantially reduced as a resul-t of azimuth
~z~
loss. Such azimuth loss La, during reproduction, as a
result of the recording of the video signal in adjacen-t
tracks with different azimuth angles, can be expressed as
follows:
sin ~AW tan ~
La = 20 log ~ tan ~ d~ ... -(5)
where ~ represents -the azimuth angle relative to a reproducing
(or recording) head 1 and recordlng track 2. It should be
appreciated from e~uation (5) that, if the relative speed of
head 1 to the tape is constant, the azimuth loss La lncreases
with decreasi.ng wavelength ~, that is, with increasing
frequency. Thus, in accordance wi.th another aspect of this
invention, the digital video signal is code converted in order
to reduce the low frequency spectrum components thereof so as
to provide an increased azimuth loss La.
In particular, a specific example will now be
illustrated or comparing the recording according to this
invention with the previously discussed recording which
provides guard bands between adjacent tracks. As shown in
Fig. 3A, the digitized video signals from two channels are
recorded in adjacent tracks 2 wlth a differential azlmuth
angle ~d between the adjacent ~racks which ls selected as 14,
and the width W of each track is selected as 60~ m wlth no
guard bands therebetween. When a reproduclng head 1 (wlth a
-16-
2~
eorrect air gap angle) traces a respective one of tracks 2 of
Fig. 3A, the cross-talk component Erom the adjacent -track is
shown by eurve C2 in Fig. 2. In comparison, Erom me~sured
values, when the track width W is selec-ted as 40~Jm, the wid-th x
of eaeh guard band is selected as 20~ m and the azimu-th ang]e
~ for all tracks is selected as 0, as shown in Fig. 3B, and when
reproducing head 1 scans one of tracks 2 in Fig. 3B, the actual
cross-talk component from the adjacent track is represen-ted
by eurve C~ in Fig. 2. I'he above curves C2 and C3 were obtained
:Eor the ease where the relative speed oE the reproducing hea~
to the tape was equal. to 25.59 m/sec.
From eurve C2 according -to this invention, for
frequencies lower than approximately 2 MHz, -the cross-talk
eomponent decreases with lncreasing frequencies up to
appro~imately 2 MHz. When the frequency is higher -than
2 MHz, the cross-talk interference between adjacent tracks
.increases with increasing frequency due to coupling between
the heads and the like. In the actual curve C3 in which the
recorded traeks have guard bands -therebetween :Eor frequencles
lower than about 200 KHz, the cross-talk curve becomes colncldent
with the theoretica1 eurve Cl. Above 200 KHz, curve C3 follows
a si.milar pattern to that of eurve C2 in wllich the eross--talk
interferenee from adjaeent reeorded traeks decreases wi-th
inereasing Erequencies up to about 2 MHz and thereafter,
increases with inereasing frequencies. Further, from a
eomparison of curves C2 and C3, i-t is seen that, for frequencies
lower than about 1 MHz, the cross-talk componen-t, in the
ease of azimuth recording according to this invention, is
greater than the cross-talk eomponent in the case oE normal
recording with guard bands, by only ~ to 6 dB. In the
frequency range greater than 1 MHz, curves C2 and C3 are
substan-tially coincident so that -the cross-talk components
are approximately equal.
It should be appreciated that the tape consumption
.is the same for normal recording with guard bands, shown in
Fig. 3B, and azimuth recording without guard bands according
-to this invention, shown in Fig. 3A. Fur-ther, -the cross-talk
interference between adjacent tracks for the two recordings
is substantially the same. However, as previously discussed,
the S/N ratio of the reproduced digital signal is propor-tional
to the square root of the track width W~ Thus, as the track W
is increased, the level of the reproduced signal, and consequently,
the S/N ratio, also increases. It should therefore be
appreciated that.the overall S/N ratio for the record:ing
according to this invention, as shown in Fig. 3A, is higher
than that for the recording shown in Fig. 3B. In par-ticular,
the S/N ratio for the recording according to this inven-tion is
greater than that for -the recording shown in Fig. 3B by an
amount 20 log ~ = 1.76 dB.
Further, upon the occurrence of a -trac]cing error by
the reproducing head, it should be appreciated that -the S/N
ratio of the reproduced signal for the recording as shown in
Fig. 3A is even higher than the aforementioned 1.76 dB over
that for the recording shown in Fig. 3B. For example, when
reproducing head 1 is displaced so as to trace two adjacent
tracks by an e~ual amount, as shown in Figs. 3A and 3B, any
deterioration of the S/N ratio for the recording in Fig. 3A
-18-
;20~
is substantially reduced as a result of azimu-th loss. However,
when reproducing from the recorded tracks shown in Fig. 3B by
reproducing head l which equally overlaps the two traclcs, the
S/N ratio is 0 dB since the amount of cross-talk interference
picked up from the non-desired track is equal tG the level o~
the signal from the track desired to be traced. It should
therefore be appreciated that the utilization of azimu-th
recordiny with no guard bands between adjacent tracks provides
a greatly improved recording over -tha-t previously proposed.
In other words, by utilizing an azimuth recording, -there is
obtained a high S/N ratio while also providing high bit density
recording.
There is, however, a limit to the value of the
azimuth angle ~. In particular, the effective recording
wavelength ~e can be expressed as follows:
~e = Acos ~ .... (6)
where ~ is the actual recording wavelength utilized. Frorn
equation (6), it should be appreciated that the effective
recording density is lowered, and consequently, the recording
~is easily effected by spacing and gap loss when -the effective
recording wavelength ~ e is small. Since -the effective recording
wavelength ~e decreases as the azimuth angle ~ increases, the
differential azimuth angle ~d be-tween adjacent tracks cannot
be selected too large. It has been ascertained by experiment
that the differential azimuth angle 3d is preferably selected in
the range of 10 to 30 in order to provide high density
reco~.ding.
--19--
~8~9
As previously discussed in regard to equation (5~,
the azimuth loss La increases as the recording frequency
increases. In :Like manner, when the recording frequency is
low, the azimuth loss La is also low. This is seen more
particularly by curve C2 in Fig. 2 which illustrates an
increase in the cross-talk interference with decreasing
frequencies below approximately 2 MHz. It should be appreciated
that the cross-talk interference between adjacent tracks
is considered as a noise signal, in addi-tion to other previously-
mentioned noise components, which results in a deteriora-tion
of the S/N ratio of the reproduced digital signal. Since the
S/N ratio for the reproduced digital signal must be greater
than 20 dB, as previously discussed, the level of the cross-
talk interference must be lower than approximately -30 dB.
Thus, for example, in the case of azimu-th recording shown
by curve C2 in Fig. 2, the level oE the cross-talk interEerence
is lower than -30 dB when the recording frequency is in
the range of approximately 1 MHz to 25 MHz. However, the
digitized video signal converted from the analog video signal
includes many components wi-th frequencies less than 1 M~z.
Thus, in accordance with another aspect oE this inven-tion,
the occurrence of low frequency signal components of the
digitized video signal is reduced so as to subs-tan-tially reduce
cross-talk interference which cannot satisfactorily be
eliminated by means of azimuth loss.
In particular, the present invention utilizes a
code conversion system in which the digitized video signal is code
-20-
\
~22~
eonverted to eliminate or at least substantially reduce such
low frequency eomponents in the digitized signal. Various
types of code conversion systems are kno~ in the ar;t. For
eY~ample, if the original digitized signal is an NRZ (non-return-
to-zero) signal (Fig. 4A), it may be code converted to, for
example, a bi-phase code signal ~Fig. ~B), a Miller code
signal (Fig. 4C) or an~l2 or modified Miller code signal (Fig.
~D), the frequency speetra of such signals being shown in
the graph of Fig. 5, respectively. In the graphical diagram
of Fig. S, r represents the bit period, fs represents -the
-transmitting frequency (that is, ~he recording bit ra-te),
and fn represents the Nyqulst frequency. It should be
appreciated that when the digitized signal is converted from
the analog signal, it is in parallel form. However, upon
reeording, the digitized signal i.s eonverted from parallel
form to serial form and the transmitting frequeney fs is the
Erequeney of the serial digitized signal. It should further
be appreeiated from Fig. 5 that the above eode eonversion
systems, that is, bi-phase, Miller and M2, reduce the low
frequeney eomponents of the digit:ized signal in comparison
to the original NRZ dlgitized signal (Fig. ~A).
In aecordanee with another eode eonversion sys-tem,
the digitized signal is eode eonverted in an 8--to-10 code
eonversion process, that is, a digitized signal comprised of
8-bit words is converted to a digitized signal comprised of
10-bit words. The broken line in Fig. 6 represents the
theoretieal frequency distribution with such 8-to-10 conversion
proeess and the solid line represents the actual frequency
distribution thereof. Pr~ferably, the block coding is such
-21-
2;~
that 28 codes whose DC levels are close ~o zero are selected
from 2~ codes of 13-bi words and arranged to have one-to-one
correspondence to the original 8-bit codes, as specifically
disclosed in de~ail in Patent Application Serial No~ 355,345
file~ July 23, 1980, having a common assignee herewith. By
means of this process, the DC level of ~he recorde~ signal is
made as close ~o zero as possible, that is; ~D" and "1" bits
alternate with each o~her as much as possible. Thus, for
example, if f5 equals 38.4 MHz, as shown in Fig. 6, the lower
cut-off frequency at which the frequency spectrum is evenly
divided in hal~ is approximately 1.3 MHz, and in the frequency
range lower than this cut-off frequency the frequency spectrum
falls sharply. In this manner, the occurrence of low
frequency components of ~he digitized slgnal, that is,
components having a frequency below 1.3 MHz which result in
the level of cross-talk interference being above -30 dB lFig.
2), is substantially reduced. ~hus, the azimuth loss La is
i~ubstantially increased so as ~o more effectively reduce cros5-
talk interference from adjacent tracks. In this manner, high
density recording is achieved while reproduced digital signals
from the recorded ~racks have a high S/N ratio.
There will now be described an apparatus according
to this invention for performing the above--described method
of recording a digitized video signal in a plurality of parallel
tracks extending obliquely on the magnetic tape without
guard bands between at least some of the adjacen~ track~ and
with the digitized video signal in some of the Darallel tracks
being recorded with an azimuth angle which is different from
~he azimuth angle in other ones of the parallel tracks.
-22-
,;'-' ''`
.
~2;2~
However, in order to facilitate an better understandi.ng of this
aspect o:E the present invention, there will Eirst be described
the conditions for digi-tal recording oE, Eor example, an NTSC
color video signal.
The NTSC system color video signal is desirably
digitized with the following conditions being established:
1. Since one frame comprises 525 lines, the
number o:E lines selected for a first (third) and a second
(Eourth) field are 262 and 263, respec-tively. In -the firs-t
field, a vertical synchronizing pulse and a horizontal
synchronizing pulse are in phase wi-th each other, and the
field in whi.ch they are out of phase is considered the
second field.
2. The number of sampled picture elemen-ts in
each horizon-tal period (H) varies with the sampling frequency
(Es) employed. Since the color sub-carrier frequency (fsc)
is 455/2 times the horizontal frequency (fH), the numbers oE
sampled picture elements in one horizontal peri.od are as shown
in the below Table 1 in the case of fs = 3fsc and in the
case of Es = 4fsc
Table 1
E
s Even line Odd line
Odd frame 682 5~3
3fsc
Even frame 633 6~2
Odd frame 910 910
sc
Even frame 910 910
-23-
Apparatus for performing the previously-described
recording arrangement according to the present invention will
hereinafter be described wi-th reference to a recording
section (Fig. 7) and a playback or reproducing section (Fig. ~)
of a digital VTR which will now be described in greater
detail. In the digital VTR, a digitiæed video signal is
recorded by a rotary head assembly (Fig. 10) in parallel
tracks extending obliquely on a magnetic -tape 3 (Fig. 11).
Since the transmitting bit ra-te oF the digital video signal
:is high, as previously discussed, four rotary heads lA, lB,
lC and lD (Fig. 9) are disposed in close proximi-ty -to each
other, and the digitized video signal of one field is
distributed -through four channels to such heads and recorded
on the magnetic tape in four parallel tracks.
Referring in detail to Fig. 7, it will be seen
that an NTSC color video signal to be recorded is applied
through an input terminal 11 to an input processor 12. The
input processor 12 comprises a clamp circuit and a synchronizing
and burst signal separator and supplies the effective or video
informatlon portion of the color video signal to an A/D
converter circuit 13. A synchronizing signal and a burs-t
signal separated from the color video signal by processor 12
are applied to a master clock generator 21 which is desirably
of PLL (phase-locked loop) construction. The master clock
generatOr 21 generates clock pulses of the sampling frequency,
for example, 4fsc or 4 times the frequency of the burst signal.
The clock pulses from generator 21 and the synchronizing
signal are applied to a control signal generator 22 which produces
various kinds of timing pulses, identifying signals (ID) for
2~'9
identifying lines, fields/ frames and tracks, and a control
signal, such as, a train of sampling pulses.
The A/D converter circuit 13 generally cornprises
a sample-and-hold circuit and an A/D conver-ter for conver-ting
each sampled output to an 8-bit code which is supplied, in
parallel form, to an interface 14. The duration or period
of one line (lH) of the NTSC color video signal is 63.5~s
and a blanking period therein is 11.1~ s. Accordingly, the
period of the effective vldeo region or portion is 52.4~ s.
When the sampling frequency is 4fsc = X2 55 fH~ where fH
is the horizontal frequency, the number of samples in one
horizontal period is 910. Further, the number of samples
in the effective video region or oortion is 768 samples, as
shown in Fig. 12. In consideration of the division oE the
video information to be recorded into four channels, the number
of effective vldeo samples is selected to be 768 per line
or horizontal period with 192 samples being assigned to each
channel. In Fig. 12, HD represents the horizontal synchronizing
signal and BS represents the burs-t slgnal.
The number of lines forming one field is 262.5H,
with a vertical synchronizing period and an equalizing pulse
period accounting for 10.5H. Since test signals VIT and VIR
are inserted in the vertical blan~;ing period, they are also
regarded as effective video signals. Thus, the number of
effective video lines in one field period is selec-ted to be 252.
In other words, an effec-tive frame is selected and may be
arranged, for example, so that the first or odd field thereof
includes video information in lines 12 263 and the second or
even field there~f includes video information in lines 274-525.
-25-
~ ~2~9
In this manner, each of the odd and even fields of each frame
includes 252 field lines of video information.
The digitized effective video region of -the
color video signal is divided by interface 1~ of the digi-tal
VTR into four channels. For example, with 768 samples per
line, data corresponding to samples (4n-~1) are assigned -to
channel A, data corresponding to samples (4n-~2) are assigned
to channel B, data corresponding to samples (~n+3) are assigned
to channel C, and data corresponding to samples (4n-~4) are
assigned to channel D. The da-ta of the four channels are
processed in the same manner and only one channel will be
described. The data in any one of the channels, for example,
channel A, is derived as a record signal for head lA after
being applied, in se~uence, to a time base compression circuit
15A, an error correcting or controi encoder 16A, a recording
processor 17A and a recording amplifier 18A. The recording
amplifiers 18A, 18B, 18C and 18D are connected bv way of a
rotary transformer (not shown) to rotary heads lA, lB, lC
and lD, respectively, disposed in close proximity to each
other.
The code arrangement of each of the recorded si.gnals
respectively provided at heads lA to lD will now be described
with re:Eerence to Fig. 14. As there shown, a sub-block SB of
the coded digiti~ed signal is composed of 105 samples (840 bits)
.in which a block synchronizing signal (SYNC) of three samples
~24 bits), an identifying (ID) and address (AD) signal of two
samples (16 bits), information data of 96 samples (768 bits)
and CRC (Cyclic Redundancy Check) code of four samples (32 bits)
-26-
~822(1 ~
are arranged one after another. The data of one line or
horizontal period of the color video signal comprises 192
samples per channel, as previously mentioned, and these
samples are divided into two sub-blocks, that is, there are
two sub-blocks for each line of each channel, with 96 samples
for each sub-block. In other words, each sub-block SB includes
data for one-eighth of a line. The block synchronizing signal
is used for identlfying the beginning of a sub-bloc]c, whereupon
the identifying and address signals, the information da-ta
and/or CRC code can be extracted. The identifying signals ID
indicate the channel (track), the frame, the field and the line
to which the information data of the sub-block belongs, and
the address signal AD represents the address of the respective
sub-block. The CRC code is used for the detection of an
e~rror in the information data of the respective sub-block.
~ Fig. 13 shows the code arrangement for one field
in one channel. In Fig. 13, each reference character
S~ 572) indicates one sub-block, with two sub-blocks
making up one block or line. Since the effective video region
of one field is comprised of 252 lines, as mentioned previously,
the data of 252 blocks (50~ sub-blocks) exist in one field.
The video information data of a particular field are
sequentially arranged in a 21 x 12 matrix form. Parity
data are also provided in connection with the horizontal
and vertical directions, respectively, of -the video information
data in the matrix. ~ore particularly, on Fig. 13, the parity
data for the horizontal direction is shown positioned in the
thirteenth column of blocks, and the parity data for the
-27-
22~
vertical direction is positioned in the twenty-second row at
the bottom. In the thirteen-th column of blocks at the
twenty-second row is disposed the vertical parity data
for the horizontal pari-ty data. The parity data for the
horizontal direction is formed in three ways by 12 sub-blocks
respectively taken out of the 12 blocks forming one row of
the matrix. In the first row, for example, parity data SB25
is formed by the modulo 2 addition:
[SBl] ~ [SB3] ~ ~SB5] ~ ...... ~ [SB23] = [SB25] .
In the above, [SBi] means only the data in the respecti.ve
sub bloclc SBi. In this case, samples belonging -to respec-tive
ones of the 12 sub-blocks are each calculated in a parallel,
8-bit form. Similarly, by the modulo 2 addition:
~ [SB2] ~ [SB4] ~ [SB6] ~ ...... ~ [SB24] = [SB26]
parity data [SB26] is formed. The parity data is similarly
formed for each of the second to twenty-first rows in the
horizontal direction. Enhancement of the error correcting
ability results from the fact that parity data is not formed
merely by the data of the 24 sub-blocks included in a row,
but is formed by the data of 12 sub-blocks positioned at
intervals of two sub-blocks in the row.
The parity data for the vertical direction is
formed by the data of 21 sub-blocks in each of the first to
thirteen columns of blocks. In the first column, parity
data [SB547] is formed by the modulo 2 addition:
[SBl] ~ [SB27] ~ ~SB53] ~ .... ..[SB521] = [SB547].
-2~-
In this case, samples belonging -to each one of the 21 sub-
blocks are calculated in a parallel 8-bit form.
Accordingly, these parity data comprise 105 samples
as is also the case with the video data sub-blocks. In
the case of transmitting the digitized signal of one field
of the above matrix arrangement (22 x 13) as a series of first,
second, third, ... twenty-second rows in sequence, since 13
blocks correspond to the length of 12H, a period oE 12 x 22 =
264H is needed for transmitting the digital signal of one
field. In other words, since the number of samples in each
sub-block SB is 105 and the number of sub-blocks per field in
each channel is 572, the number of samples per channel for
each Eield ls 105 x 572 = 60,060 samples. Fur-ther, since -there
are 4 channels and 910 samples per line, the number of horizontal
periods needed for transmitting the video signal of one
fteld is (60,060 x 4)/910 = 264H.
Incidentally, if the VTR is of the C-format
type, and thus employs an au~iliary head for recording and
reproduci~g one part of the vertical blanking period in one
field, then a duration of only about 250H can be recorded with
a video head. In accordance with the present invention, a
duration of 246H, leaving a margin of several H's, has to be
recorded in each track, that is, the period of 264H of data
to be transmitted is time-base-compressed (with a compression
ratio Rt of 41/44) to a period a duration of 246H. Further,
a pre-a~ble signal and post-amble signal, each having the
transmitting bi-t fre~uency, are inserted at the beginning and
the terminating end of the record signal of one field having
-29-
t~
zz~
the period of 264H. ^~
The time base compression circuit 15 in Fig. 1
compresses the video data with the above-noted compression
ratio 41/44 and provides a data blanking period in which
the block synchronizing signal, the identifying and address
signals and the C~C code are inserted for each sub-block of
video data of 96 samples, and at the same time, sets up data
blanking periods in which the blocks of the parity data are
inserted. The parity data for the horizontal and vertical
directions and the CRC code of each sub-block are generated by
the error control encoder 16. The block synchronizing signal
(SYNC) and the identiEying ~ID) and address (AD) signals are
added to the video data in the recording processor 17. The
address signal AD represents the previously-noted number (i)
of the sub-block. Further, in the recording processor 17
th,ere is provided an encoder of the block coding type which
converts the number of bits of one sample from 8 to 10, and a
parallel-to-serial converter for serialiæing the parallel 10-bit
code. As disclosed in detail in the aforementioned
Patent Application Serial No. 365,345, filed July 23, 1980
and having a common assignee herewith, the block coding is such
that 28 codes whose DC levels are close to ero are selected
from 21 codes of 10-bit words and arranged to have one-to--one
correspondence to the original 8-bit codes. By means oE the
oregoing, the DC level of the recorded signal is made as
close to zero as possible, that is, "0" and "1" alternate
with each other as much as possible. Such block coding is
employed for preventing degradation of the transmitting
waveform on the playback side by substantial DC free
transmission. It may also be possible to achieve the same
-30-
.. ...
results by employing a scramble system utilizing the so-called
M-sequence which is substantially random in place of the block
coding. It should be appreciated that by means of such code
conversioll, the low frequency signal componen~s of the digital
video signal are substantially reduced so that, for example,
only signal components with frequencies higher than
approximately 1.3 MHz are produced, as previously described
in regard to Fig. 6. The recording processor 17 converts
the code converted digital signal from parallel to serial
form and then transmits the sub-blocks sequentially to the
respective heads. In the case where each sample comprises
8 bits, the transmitting bit rate per channel, after
converting the above 8-bit code to the 10-bi~ code, is-as
follows:
(4fsc) x 8 x l4 x 4~4 x 18 ~ 38.4 Mb/sec.
This, of course, is the previously-described frequency f5 in
Fig. 6.
The sexially arranged digital signals in each
channel are respectively supplied through a recording amplifier
18 to respective rotary maynetic heads lA to lD, which are
arranged as shown in Figs. 9 and 10. In particular, each of
heads lA to lD has a height selected equal to the track width W.
Further, heads lA and lC are mounted on a rotary drum 5 and
aligned in the vertical direction~ith a distance W therebetween,
and the other heads lB and lD ar~ also mounted on rotary drum
5 and al~gned in the vertical direction with a distance W there-
between. Heads lA to lD are arrangedin close proximity-to one
-31-
~ ~Z~19
another so that, for example, head lB (lC) is positloned
in the vertical direc-tion between heads lA (lB) and lC (lD).
Further, heads lA and lC are selected to have the same azimu-th
angle ~/2, for example,7 in one direction, and heads lB and lD
are selected to have the same azimu-th angle ~/2, for
example, 7 in the direction opposite -to -tha-t of heads lA and
lC. In this manner, the differential azimuth angle ~d between
adjacent tracks is 1~.
Ileads lA to lD are rota-ted together with rotary
drum 5 .in synchronism with the color video signal a-t the
:Ee.ild frequency, and a magnetic tape 3 con-tacts the peripheral
surfaces of heads lA to lD and rotary drum 5 over an angular
range of abut 360 in a slant omega (JQ) configuration, and
the tape is driven at a constant speed. Accordingly, as shown
in Fig. 11, the digitized signals from channels A to D are
respectively recorded on tape 3 by heads lA to lD in tracks 2A
to 2D, respectively, each track corresponding to one field.
A control trac]c ~ is also formed at the lower longitudinal edge
of tape 3. In this case, the distance W between respective
ones of the heads lA to lD is equal to the track width W, so
that adjacent tracks 2A to 2D contact each other a-t the
longitudinal edges thereof without any guard bands therebetween.
Further, if the rotary radius of each of heads lA -to lD and -the
speed of tape 3 are suitably selected, track 2A of each field
may contact track 2D of the following field at the respective
longitudinal edges thereof, as shown in Fig. 11. Further,
since the azimuth angles of the heads are alternately opposed
to one another, the azimuth angles of tracks 2A to 2D are also
alternately opposed to each other so as to minimize cross-talk
-32-
2~
i
interference between adjacent tracks.
In the reproducing or playback por-tlon of -the
digi-tal VTR, as shown in Fig. 8, the four channels o;E
reproduced signals are derived from reproducing heads lA to
lD which scan tracks 2A -to 2D, respectively, corresponding
thereto, and are applied through playback amplifiers 31A -to
31D to respective waveform shaping circuits (no-t shown).
Each of the waveform shaping circuits ~includes a playback
equallzer Eor increasing the high-frequency component of the
reproduced signal and shapes the reproduced signal to a clear
pulse signal. Further, each waveform shaping circuit ex-trac-ts
a reproducing bit clock synchronized with the pre-amble signal
and supplies the reproducing bit clock -to respective playback
processors 32A to 32D -together with the data. In each o-f
playback processors 3~A to 32D, the serial data is converted
to parallel form, the block synchronizing signal is extracted,
the data is separated from the block synchronizing signal and
from the ID, AD and CRC codes or signals, and further, block
decoding or 10-bit -to 8-bit conversion is performed. The
resulting data is applied to time base correctors 33A to 33D,
respectively, in which time base errors (or axis -Eluc-tuations)
~re removed Erom the data. Each of time base correctors 33A
to 33D iS provided with, for example, four memories, in which
reproduced data are sequentially written by clock pulses
synchronized with the reproduced data, and the data are
sequentially read out from the memories, by reference clock
pulses. When the reading operation is likely to get ahead
of the writing operation, the memory from which the data has
just been read is read again.
-33-
The data of each channel is provided from the
respective ones of the time base correctors 33A to 33D to
respective error correcting decoders 34A to 34D. Each
error correcting decoder 34A to 34D includes error detecting
and correcting circuits using CRC, horizontal and vertical
parities~, a field memory and so on. In particular, each
error correcting decoder includes a field memory, and da-ta is
written in-to each field memory a-t every sub-bloc]~ SB in
response to, for example, the respective address signal AD -there-
of. At this time, any error in the data is corrected for every
sub~-bloc]c SB of information by -the CRC code and horizontal
and vertical parity da-ta. If the error is too great to be
corrected by the CRC code and parity data, -the writing in
of data of that sub-block SB in the field memory is not
performed, and instead, data from the previous field is read
out again.
The data from error correcting decoders 34A to 34D
is supplied to respective time base expander circuits 35A to
35D, which returns the data to the original transmitting rate
and then applies the data to a common interface 36. I'he
i.nterface 36 serves to return the reproduced data of the
four channels into a single channel which includes a D/A
converter circuit 37 for conversion oE the data into analog
Eorm.
The output from D/A converter circuit 37 is
applied to an output processor 38, from which a reproduced
color video signal is provided at an output terminal 39. An
external reference signal may be supplied to a mas-ter clock
generator (not shown), from which clock pulses and a reference
-34-
8~
sync..ronizing signal are provided ~o a control signal generator
(not shown). The cvn~rol signal genera~or provides control
signals synchronized with the external reference signal, such
as, varlous ~iming pulses, identifying signals ~or the line~
field and frame, and sample clock pulses. In the reproducing
section,~ ~he pxocessing of the signals from heads lA to lD
to the input sides of time base correctors 33A to 33D is
timed by ~he clock pulse extracted from the reproduced data,
whereas.the processing of the signals from the output sides
of the time base correctors 33A to 33D to ~he output terminal
39 is timed by ~he clock pulse from the master clock generakor.
The above recording and reproducing sections shown in Figs~ 7
and 8 are d~scl.osed more particularly in Patent Applicat~on
Serial No. 362,045, filed October 9, 1980, having a common assignee
herewithO
It should be appreciated that th~ method and apparatus
for recording a video signal according to this invention,. as
described above, provides that a digi.tized video signal is
recorded in parallel tracks with a high recording density in
such a manner so as to reduce the tape consumption, resulting
in longer periods of recording being possible with a given
leng~h of magnetic tape. Further, because the digitized
video signal is recorded in parallel tracks without guard bands
between adjacent tracks, the tracks are wider so as to increase
the S/N ratio. Further, since more information can be recorded
on the wider tracks, the recording density on the ta~e can be increased. In
~35-
addition, because of -the more efficient utilization of tape
by providing wider tracks, there exists a larger tolerance
for any tracking er.ror during the reproduction operation.
Since a multi-head apparatus is provided, i-t also may be
possible during the reproduction opera-tion to de-tect and
correct any tracking error by means of the phase difference
between the outputs from, for example, heads lA and lB. It
should be appreciated that al.though the presen-t inven-tion has
been described in the normal mode of operation, o-ther special
reproducing modes may be utilized with this inven-tion, for
example, a search mode,in which -the heads scan the tracks
when the relative speed between the heads and the tracks is
greater -than that for normal recording. Thus, by iden-tifying
the channel with the identifying qignal ID, reproduction during
such search mode can be carried out even when, for e~ample,
head lA, scans track 2C.
~ Although one embodiment of the present invention
¦ has been described above, it should be appreciated that other
embodiments within the scope of this inven-tion can be provided.
E`or example, although the digitized video signal was d:ivided
into Eour channels and the signal of one field of video inEorma-
tion was recorded in four tracks 2A to 2D, the digitized video
si~nal may be divided into an odd number of channels, for
example, three channels. In such case, -three recording
(reproducing) heads are provided, each associa-ted with one
of the channels, so that the three heads simultaneously
scan tracks 2A to 2C, respectively, to record one field of
video information, as shown in FigO 15. In such case, a
~ guard band GB is form~d after every -three sequential tracks
-36-
I
~22~9
2A to 2C so as to separate each field of video informa-tion on
tape 3. However, it shoilld be appreciated -that the azimuth
angles in adjacent tracks without guard bands Gs therebetween
are different. For example, it may be sufficient for each
center track 2B to have the digitized video signal recorded
therein with a first azimuth angle and the remaining two
tracks 2A and 2C having a digitized video signal recorded
thereirl with a second azimuth angle which is different from
the first azimuth angle.
E`urther, although it is preferable that the low
frequency signal components of the digitized video signal
be attenuated by the aforementioned i3-to-10 conversion system,
it may be the case that the cross-talk interference between
tracks at the outputs of reproduci.ng amplifiers 31A to 31D
is smaller than a predetermined vaiue. In such case, an NRZ
re,cordi.n~ partial response detecting gystem, which will
attenuate low frequency signal components in the reproducing
section, may be sufficlent to provide a sa-tisfacotry S/N ratio.
Further, it should be appreciated that although
a magnetic tape 3 has been utilized in -the preferrecl embodiment
of this invention, it may be possible to use a magne-tic disc,
magnetic drum or the like.
Having described specific preferred embodimen-ts
of the invention with reference to the accompanying drawings,
it is to be understood that the invention is not limited to
those precise embodiments, and that various changes and
modifications may be effected therein by one skillecl in -the
art without departing from the scope or spirit of -the invention
as defined in the appended claims.
