Note: Descriptions are shown in the official language in which they were submitted.
. PIIN 820iC
RJ
~t~r~ ~7~9-76
"Apparatus for reading a record carrier~'.
The invention rolates to an apparatus for
reading a record carricr on which a television sig-
nal is reoorded, which signal comprises horizontal
I . . syn~h~onizing pulses and burst signals which are
¦ 5 . coupled to these horizontal synchronizing pulses,
¦ which burst signals`consist of a number o~ periods
o~ a~carrier wave ~ith a frequency which is an in-
; tegral multiple of half the line frequency, which
~ - read apparat~ls is provided ~ith a time error cor-
'
rection device for correcting time errors in the
. . read-out television signal and a time error detec-
. ~ ~ tor for detecting said time errors and supplying a
.. corresponding control signal to the time error cor-
.
. rection device 9 whioh time error detector comprises
a kéying signal generator which is triggered by the
¦ . . horizontal synchronizing pulses, whlch generator
supplies a keying signal which is delayed by a first
. time inter~al relativ0 to said horizontal synchro-
- .
nizing pulse, said first time interval being such
that the beginning of the keying signal each time
: lies within the time interval occupied by the burst
: ~ :
signal, and furthermore a zero passage detector for
.
detecting:the first zero passage of the burst sig-
nal which appears a~ter the be~inning of said keyi.ng
2 - ~
.
PllN 8201C
27.9.76
2~3
signal.
In this respect it is to be noted that a
re¢ord carrier generally contains sùch a television
signaL in coded form, i.e. that this television sig-
S nal is coded iIl a specific manner be~cre it is re-
cordcd. In accordance with a frequently used coding
the complete television signal) for example a standard
PAL or NTSC colour television signal, is frequency
modulated on a carrier wave. In other coding systems
the luminance and chrominance signals are added to
separate carrier ~aves. The coding to which the te].e-
vision signal has been subjected during recording is
irrelevant for the present invention, provided that
this television signal still exhibits the specified
composition with-horizontal synchronizing pulses and
associated burst sig~als after the complementary de~
codingO
Signals which are read from a record carrier
generally sxhibit time errors. When reading record 3
.20 carriers in the form of a tape, these time errors
are inter alia caused by variations in the trans-
.. . .
port speed and stre~ch of such record carriers. When
reading disc-shaped record carriers these time errors
may bs caused partly by variations in the transport
speed, (in this case the speed of rotation) of the
record carrier and partly by eccentricity of the
; drive point relative to the centre of the record
carrier.
,
.
.. .
Pl-IN 8201C
27.9~76
These time errors are very disturbing when
reproducing a television signal which is recorded on
a record carrier in particular in -the case of a colour
tele~ision signal in which two colour ColJlpollentS are
quadrature-modulated on one and the same colour car-
rier wave~ In such colour television systems said time
errors produce very annoying colour hue variations~
which are not acceptable.
Apparatus for reading such record carriers
therefore usually include a tim~ error correc$ion sys-
tem, with which these time errors can be corrected as
far as possible. In such read apparatus both electro-
mechanical ser~o systems and fully electronic systems
may be used as time error correction systems~ For
~, 15 example in apparatus for reading record carriers in
the form of a tape it ls common practice to c~ntrol
both the transport speed of the record carrier and
the speed with which the read head moves along the
record carrier in accordance with the time errors
measured. In the case of apparatus for reading a
disc-shaped record carrier the same is effected with
respect to the speed ~ rotation of the disc-shaped
record carrier. Moreover, such read apparatus ge-
nerally include an addi~ional correction system in
the scanning~unit, In the case of a mechanical
scanning system this additional correction system
~or example controls the po~ition of the scanning
'
- ~
P}IN 8201C
27.9.76
Zfi~
head in the longitudinal direction of the track on
the record carrier. In the case o~ an optical reacl
system, the in~ormation on the record carrior is read
with the aid of a beam of radiation and the position
of the scanning spot produced on the record carrier
by said beam is controlled in the longitudinal direc-
tion o~ the track. Electronic time error correction sys-
. tems may emplo~ variable delay networks such as bucket
brigades, CCD~s (charge-coupled devices) and the
like.
In all these time error correction devices
it is of great importance that the time errors con-
tained in the read-out television signal can be
measured with high ac¢uracyO In addition, it is
: 15 desirable that suoh a time error correction device
., .
-: has a fairly wide control range, for which purpose
the time errors should be measured over a substan-
~ tially wide range~
- In "I.E.E.E. Transactions on~Broadcasting",
Vol. BC-17, No. 1, March 1971, page 35, a time error
-: :
detector device for an NTSC colour television signal
is desoribed which meets these two requirements. In
this device use is made of the colour burst signal
. ; on the backporch of the her~R~Dtal synchronizing
~:~ ; 25 pulse in this NTSC colour televisiol1 signal. This
:
- colour burst signal consists o~ a number of periods
of a c~arrler wave of a frequency whlch is an odd
,;; ~:
: - 5
,
,
P~IN 8201C
27.9.7
:~t~2~3
integral multiple of half the line frequency, said
burst signal having a fixed posi-tion relative to the
correspondlng horizontal svnchroniz:ing pulse. A key-
ing signal generator which is triggcred by this ho~
rizontal synchroni~ing puls0 produces a koying sig~
nal, which lies within the time interval occupiecl
by the co:Lour burst signal. By detecting the first
colour burst zero passage appearing after the begin-
ning of this keying signal a pilot signal of line
frequency is obtained from which the desired control
signal for the time error correction device is deriv-
ed by comparison with a reference signal of line fre-
quency. As a zero passage of the colour burst signal
can be detected in a very accurate ~anner, the time
error meaaurement thus becomes highly accurate.
Moreovar, ~inoe in fact use is made of a pilot tone
of line frequency~ the time error measuring range,
- and thus the range of the time error correction, cor-
responds to one line period of the television signal.
For the time error detection described above
it is essential that the beginning of the keying sig-
nal within the time interval of the burst signal is
well defined, in order that always the same zero
passage of said burst signal ultimately functions
as the pilot signal. However, the relative positions
of the horizontal synchronizing puls0, specifically
its leading 0dge, and said burst signal ar0 not ac- ;
., ~, ~ .
~ 6
.
.
PIIN 8201C
2'7,9.76
Z~
curately defirled. As a resul-t of this, the beginning
of the keying signal, which is derived from this ho-
rizontal synchronizing pulse, is neither accurately
defined. This may lead to an erroneous time error
n~easurelnent in that, owing to -the change of the zero
passage of the burs-t si~nal which acts as a pilot
signal, a time error is indicated without the ac-
tual occurrence of a time error in the read-out
signal.
It is an object of the invention to avoid
this problem and the apparatus according to the in- t
vention is therefore characterized in that the time
error detector comprises a measuring circuit for
measuring the time interval between the beginning
of this keying signal and the next zero passage of
the burst signal which is detec-ted and supplying a
correspoAding correction signal, and that the keying
s signal generator is adapted to supply a keying sig- !
nal which relative to the horizontal synchronizing 3
pulse has a delay equal to a first time interval
which is variable depending on the correc-tion sig-
nal.
The step in accordance with the lnvention
ensures that the beginning of the keying signal is
always situated halfway between two consecutive
zero passages of the burst signal regardless of~
changes in the posi-tion of the horizoAtal synchro-
',
- 7
PHN 8201C
~ jt~ ~ : 27.9.76
nizing pulse relati~e to the burst signal r thus po-
sitively ènsuring that always the same zero passage
of said burst signal~functions as pilot signal for
the time error d0tection.
, The invention is by no means limited to
the use of the colour burst signal as s-tated here~
inbefore with respect to the NTSC colour television
signal. As an alternative an additional burst~slg-
nal may be added to the television ~ignal to be re-
corded in an appropriata time inter~al during each
line period. An example of this is the insertion of
a burst signal during the time interval of the ho-
rizontal synchronizing pulse 9 resulting in a burst
signal which i5 superimposed on said horizontal
~, - - 15 synchronizing pulse, which possibility is describ-
ed in "Consumer Electronics~', 5-1-76, pages 54 and
further. The use of such an additional burst signal
,
is of particular importance when utilizing ~ PAL-
standard colour television signal, because in this
case the colour burst signal itself cannot be used
.
~ for the, describ0d time error detection owing to
.
the standard coupling with the llne frequency.
, Although~the invention is by no means
limited to such a system, the~invention will be
described hereinafter on the basis of an optical
, read system ~or disc-shaped record carriers. In the
drawing: ~ ~
~ ' , - - ' ,
~ 8
'' - `: :
'
PHN 8201C
27.9 .76
Zj~;t~;3
Fig. 1 shows such an optical read system,
~ig. 2 schema-tically shows t~e known time
error measuring system, and
Fig. 3 shows the signals associated with this
system.
Fig. 4 shows the time error measuring system
in accordance with the invention, and
Fig. 5 sho~s the associated signalsO
Fig. 6 shows a practical embodiment of the
~eying signal generator utilized in the apparatus
- in accordanoe with the invention.
Fig. 7 is an extension of the measuring
system, and
Fig. 8 represents an additional burst sig-
nal on the line synchronizing pulse.
The read apparatus shown in Fig. 1 is suit-
able-for reading a disc-shaped rccord carrier 1, on
whose upper surface tracks are formed (e.g. as turns
of a spiral) in which the information is recorded
a relief pattern. This track form and method of cod-
ing on the record carrier is irrelevant for the
principle of the invention~ The disc-shaped record
carrier 1 is rotated in a direction V1 with the aid
of a motor 2 with a disc-supporting spindle 3, which
proJects through a central opening 4 of the record
ca~rier 1~
~he optical system for reading the record
_ 9
PHN 8201C
27,9,76
6'~
carrier 1 is accommodated in a housing 5. This optical
system first o~ all cosnprises a radiation source 6.
This radiaticn source ~mits a radiat:ion bearn a1 which
via a semi-transparent mirrcr 7 impinges on a mirror
8 and is reflected by this mirror 8 as a radiation
beam a2. This radiation beam a2 is reflected as a
radiation beam a3 by a mirror 9, which beam is fo-
cussed by a lens 10 into the scanning spot P on the
upper surface of the record carrier 1. This upper
surface of the record carrier 1 is reflecting, so
that a radiation beam al~ (which is modulated hy the
information) is reflected and via the lens 10 impinges
on the mirror 9 is reflected as a radiation beam a5
and is then reflected by the mirror 8 as a radiation
beam a6. This radiation beam a6 impinges on the semi-
~ transparent mirror 7 so $hat a part of this radiation
bcam is re~lected as a radiation beam a7 which is in-
cident on a read detector 11. This schematically re-
presented read detector 11 detects the information
contained in the radiation beam a7 and supplies a
corresponding electrical signal to an output ter-
minal 12.
This signal at the output terminal 12 is
applied to a decoding de~ice 18, which converts the
applied television signal (which is coded in a spe-
ciflc manner) into a standard televislon signal
- which iB available at a terminal 19. It is obvious
. , .
- .
- 10
.
PHN 8201C
27.9.76
~ Z~7~
that the design of this decoding device is determined
by the coding of the television signal used during
recording on the record carrierO If, for example,
a composite system is employed,~in which the complete
standard television signal (i~e. the complete combi-
nation of lum.inance and chrominance signals~ is added
to a carrier wave as frequency modulation, this de~
coding device employs a frequency d~modulator. The
coding method is irrelevant for the principle of the
invention.
In order to ensure that th~ scanning spot P
is always incident on the information track on the
record oarrier, a radial tracking control system is
provided with which the radial position of the scan-
ning spot is controlled. This control system first
of all comprises a measuring detector for measuring
the radial position of the scanning spot P. For the
sako of simplieity it is assumed that this measur-
ing detector-is included in the read detector 11
a~d that it supplies a control signal to a terminal
13. Examples of systems with which this radia1 po-
sition of the scanning spot can be measured can be
found in the said United States Patent Specifica-tion
No. 3,381,086, United States Patent Specification
No. 3,876,84~ (PHN. 6292), and Canadian Patent
Spocification No. ~57,o67 (PHN.5503). rnis control
signal at the terminal 13 is applied to a drive
.
"'
. , .
- - - . - - . . ~ . . ~ . :
P~IN 8201C
27.9.76
IZf~
means 15 by which the mirror 8 can be pivoted in the
direction V2 about the axis 14. By pivoting this mir-
ror 8 the di.rection of the beam of radiation a1 is
changed and thus the radial position o~ the scanning
spot P. From the avera~e angular position of this
mirror 8. a second control signal i5 derived for a
drive means 16 by which the housing 5 can be moved
in a radial direction V3.
Owing to the variations in the speed o.~ the
drive motor 2 and/or eccentricity o~ the opening 4
relative to.the centre of the record carrier 1, the
read-out television signal exhibits time errors,
which are dis-turbing, particularly with respect to
the colour reproduction. In order to correct these
.15 time errors a time error correction system is includ-
ed in the optical read system. This ti.me error cor-
rection is achieved with the aid o~ the mirro 9
which is rotatable by means of a drive element 17
in a direction V4. By pivoting the mirror 9 in this
: 20 direction the scanning spot P is moved in the longi-
tudinal direction o~ the in~ormation track on the
. record carrler 1, so that said time error -can be
corrected.
The control signal required ~or said drive
element 17 i9 obtained with the aid of a time error
detector 25. The decoded television signal is applied
to this time error detector. The time error detector
.
.
PI-IN 820iC
27.9.76
~ 7~
25 first of all comprises a device 20 for extracting
a pilot signal suitable for time error measurement
from said decoded television signal. The circuit ar-
rangement of this device 20 will be described later.
The pilot signal, which has been extracted by the
device 20 and whose frequoncy and phase represent
the time errors, is applied to a phase compa-rator
circuit 21, which also receives a reference signal
from an oscillator 22. The measured phase difference
between the two signals is then a measure of the time
errors and is used to ~ tain the desired control
signal for the drive element 17 via a control am-
plifier 23. The Figure also shows a possibility o.f
deriving the control signal f`or the drive motor 2
for the record carrier 1 Prom said time error de-
tector 25 via a control amplifier 24. It is evident
that any other pilot signal contained in the .tele- ¦
vision signal may also be used for thls purpose,
because the two control s~stems need not comply
with the same requirements.
The previously mentioned known form of the
devicc 20 for extracting a suitable pilot signal is
schematioally shown in Fig. 2, whilst Flg. 3 shows
the associated signals~ In order to clarify the
operation of the device of Fig. 2 reference is
first of all made to Fig. 3a. Fig.~3a shows a ho-
rizontal synchroni~,ing pulse S followed by the
- 13
,
PHN 8~01C
27.9.76
2~3
colour burst signal B in accordance with the NTSC-
standard on an enlarged scal~. To simplify the draw-
ing the time scale has been in-terrupted within the
hori~ontal synchronizing pulse. The ~requency o~ the
colour burst signal B is an odd integral mul-tiple of'
hal~ the line frequency. This means that exactly
one line period a~ter a specific ~ero passage of
-this colour burst signal another zero passage of
' the colou~ burst signal corresponding to the next
line peri.od appears. This is utilized in the de-
vice of Fig. 2.
This device has been adapted to detect al-
ways the same zero passage of the colour burst sig-
- nal, so that a line-sequential pilat signal is ob-
tained, which on the one hand results in a wide con-
trol ran,ge of the time error correction device and
on the other hand enables accurate time error measure-
.
ments to be made. This accuracy result from the fact
that the edges of the colour burst signal are very
steep, thus enabling the relevant zero passage to
be detected in an accurate manner, whilst moreover
the disturbing influence of noise which is super-
imposed on the read-out television signal can sub-
stantially be reduced in a simple manner with this
zero passage detection 'by previously passing t~is
,
- burst signal through a narrow-band ba,nd-pass fil-'
' ter.
.
. PHN 82010
~,7~3 27,9.76
In order to ensure tha-t always the same zero
passage o~ the colour burst signal is detected, -tho
device 22 of Fig. 2 comprises a pulse gcnerator 38,
This pulse generator 38 is triggered by -the horizon-
tal synchronizing pulse, specifioa:l.ly by the leading
edge thereo~ 9 between -the horlzonta.l blanking level
VB and the peak level VT, which synchronizing pulse
is extracted from -the decoded -television signal.
For this purpose the device 20 includes a low-pass
1-0 filter 36, which is connected to the input terminal
31 and to a.sync separator 37 which is connec-ted to
this low pass filter 36, which separator in known
manner detects the horizontal synchronizing pulses
S, for example with a threshold value detector with
a threshold value VD, and supplies pulses (see Fig.
3c) which coincide with the synchronizing pulses
to the keying signal generator 38. This keying sig-
nal generator 38 comprises a first monostable multi-
vibrator 39 which is triggered by these horizontal
synchronizing pulses and in response thereto supplies
pulses T(see ~ig. 3d)~ the duration of -these pulses
T being selectsd so that the trailing edge of these
pulses T appsars within th.e timo interval occupied
. by the burst signalO This pulse T is applied to a
second monostable multivi~rator 40, which is trig-
gered by the trailing edge of this pulse T and in
response thereto supplies a pulse W (see Fig. 3e).
- 1S
,
_ . :
PIlN 8201C
27.9,7G
~7~3
This pulse W for example has as a pulse width of 140
nanoseconds, ~.e. a half period of the colour burst
signa:L~
The output o:~ the pulse generator 38, which
supplies this pulse 1~, is connected to a first input
45 of a gate circuit l~1. Ta a ~econd input l~l~ short
pulses (see Fig. 3b~ are applied, which represen-t
the ~ero passages of the burst signal B. For this
purpose the device 20 includes a low-pass filter
: 10 32 by means of which the frequency band in which
the frequency of said burst signal is situated is
extracted from the television signal applied to the
input terminal 31. The extracted signal is amplifie~
and limited with the aid of a limiter circuit 33 in
order to obtain a square wave signal and the zero
passages of this square wave signal are deLected
. with the aid of a de-tector 34, The detector 34
operates in an absolute manner in the case of an
NTSC colour television signal, i,e, the detector
34 supplies an output pulse upon each zero passage
of the burst signal, regardless of the sign of the
slope, This is necessary in the case of an NTSC
colour television signal because the burst signals
B and Bt in two consecutive line periods are in
phase opposition (see Fig, 3a) o~ing to the fact
that~the chrominance carrier ~requency is locked
-to half the line frequency~
.
- 16
~; , '
.
PHN 8201C
27.9.7G
~ 3
The gate circuit 41 (perI'orms an AND-function
for the signals at its inputs ll4 and 45, which means
that the output pulse of thls gate circui-t 1-~1 f;~
(see Fig. 3b) always corresponds to that pulse from
the æero passage detector 34 wh:ich appears within
the time interval of the pulse W from the pulse ge-
nerator 38. As a resul-t, a pulse is generated line-
sequentially (Fig. 3f) which can be utilized as a
pilot signal for time error measurement. If desired,
10 ~ this pulse may also be applied to a monos-table multi-
- ; Yibrator 42, which derives pulses with a specific
pulse width therefrom, which are th0n a~ailable at
the output terminal 43 and may be applied to the
compara-tor circuit 21 of Fig. 1 as a pilot signal.
If the pulse width selected for the pulses supplied
by the monostable multivibrator 42 is greater tllan
- the pulse width of the pulses W frorn the pulse ge-
~ nerator, substan-tially any pulse width may be used
; for these pulses W, because then only the first
pulse from the zero passage detector 34 will re-
sult in an outpu-t pulse of the multivibrator 42
during a pulse W.i
In order to increase the reliability of
operation of the device it is useful to include a
detectQr 35 which is connec~ed to the low-pass
filter 32 and which detects the presence of a
colour burs-t signal. This detector conscquently
~ ~ .
- 17
,
PllN 8201C
27.9.76
Y~ 3
supplies an output pulse as soon as and as long as
this colour burst signal is present. This output
pulse o:f the colour burst detector ls also applied
to the gate circuit 1~1 (input 1~6), so 1;ha-t this gate
circui-t can only supply an output pulse in the pre-
sence of a colour burst signal. This additional cir-
cuitry caters for the fac-t that during the vertical
flyback period of the television signal a nwllber of
picture lines contain no colour burst signals. Should
the zero passage detector 34 supply a pulse within
the pulse W, despite the absence of a colour burst
signal (for example owing to the occurrence of noise)
this does not give rise to an (erroneous~ pulse at
the output of the gate circuit.
The effect of -this ~dditional circuitry be-
comes more apparent ~len taking into consideration
that in the most frequently used version o~ the com-
parator circuit 21 a "hold effect" occllrs if no pul-
ses are supplied by the pilot tone separator. This
comparator circuit 21 generally operates with a
; sawtooth-shaped re~erence signal from the oscilla-
tor 22, which signal is then sampled at instants
which are defined by the pulses of the pilot tone
sepàrator, This sampled value is subsequently re-
tained until the next sampling oocurs. As long as
the pilot tone separator 20 supplies no pulses, i.e.
during a portion of the vertical flyback period, the
: , .
P~IN 820lC
27.9.76
'YZ~
last sample value is retained as a control signal for
the time error correction system.
However, an erro~eous pulse from the pilot
tone separator during the vertical flybacl~ perlod
would control this time error correction system in
a completely incorrect mannor. It is true that af-
ter the occurrence of ~some picture lines having
colour burs-t signals, this time error correction
system will be pulled in again, but as this is ob-
viously involves a delayed response, it could still
give rise to disturbances in the displayed picture.
In order to minimize the time during the
vertical flyback period that no measuring signal
is available it is of course also possible to add
additional burst signals to the television signal
; to be recorded during this vertical flyback period.
A problem associated with the prcviously
described method of generating a suitable pilot
- signal for time error measurements is caused by the
faot that the position of the horiæon-tal synchroniz-
ing pulse relative to the burst signal is not accu-
- rately defined. This has various causes. First oP
'
all, the accuracy with which the horizontal syn~
chronizing ~ulses are generated prior to recording
of the -television signal is liniited. Furthermore,
as previously stated, the accuracy with which these
- horlzontal synchronizing pulses can be detected dur-
- .
'~
PIIN 8201C
27.9,76
~Z~3
ing reproduction is limited and in this respect the
susceptibility to noise plays a part. Furthermore,
-the phase relation bet~een the hor:izontal synchro-
nizing pulses and tlle colour burst signal is dis-
turbed, since th:is b~rs-t signal has been passed
through a band-pass filter.
As the keying signal 1~ is directly derived
from the detected horizontal synchronizing pu]se
this means that the position of this keying signal
within the time interval of the burst signal is
- neither accurately defined. In the ideal case the
instant at which said keying signal W starts, i.e.
the position of the leading edge, ls situated exactly
halfway between two consecutive zero passages of the
burst signal. If this keying signal W is now shifted
; owing to the inaccuracy of the position of the hori-
;
zontal synchronizing pulse, it may for example happen
that the leading edge of this keying signal does no~
appear until after the desired ~ero passage of the
burst signal, as is shown in Fig. 3h~ In that case
the next zero passage of the burst signal is then
utilized for generating the pilot signal (see Fig.
31). This change from the one zero passage to the
~: next zero passage for the generation of the pilot
signal manifests itself in the tims error detection
; as a time ~rror of 140 nanoseconds, namely -the in-
t0rval between these two zero passages, thus result-
_ 20
PHN 8201C
2~.9.76
q;~73
ing in an incorrect time error measurernent and conse-
quently in an incorrect timo error correction.
This problem is avoided owing to the step
in accordance with the invention. The step in ac-
cordance with the inventioIl main:Ly affects the ci.r-
cuit arrangement of th~ Iseying signal generator 38.
In order to explain the step in acoordance with the
invention ~ig. 4 schematically sho~s the circuit ar~
rangoment of the pulse generator 38 together with
the ga-te circuit 419 which circuits form part of
the device of Fig. 2.
The pulse generator 38 sho~n in Fig. 4 oom-
prises a monostable niultivibrator 51 to whose input
54 the extracted horizontal synchron~zing pulses
supplied by ~he sync separator 37 are applied
; ~ (FigQ 2). This monostable multivibrator 51 supplies
~a pulse T'which is applied to a set input 55 of a
, : multivibrator 52, whose output is connected to the
input 45 of the gate circuit 41. The pulse from the :!
. zero passage detector 34 is applied to the input 4l~
o~ the gate circuit 41 so as to be keyed out with the
: aid of~said gate circuit, after which it is applied
to the reset Input 56 of` the multi~i'brator 52, as
the case may be after'processing by the monostable
,~ 25 multivibrator 42. Howcver, the pulse width of the
: ,
pulse T which is supplied by the monostable multi-
- vIbrator 51 iS~ not cDnstant but is ~ariable depend-
`
: ~ ~ ing on a correction signal ~hich is fed to a control
. iDpUt S7 o~ this monostable multivibrator 51.
, - 21 - '~
. . .
~. ' -- ' -'. . '. '. . '; . '' , ' '-, :.
P~IN 8201C
27.9.76
~7~
The operation of the circuit will now be
described with reference to Fig. 5. Fig. 5a shows a
number of pulses which are supplied by the zero pas-
sage detec-tor 34 and which thus represent the zero
passages o~ the burst signal. Fig. 5b shows a pulse
T supplied b~ the monostable multivibrator 51, whose
leading edge again coincides with the leading edge
of the horizontal synchronizi.ng pulse and whose
trailing edge of situated within the time interval
occupied by the burst signal. In the ideal situation
the trailiDg edge of` this pulse T should occur at
an instant t exactly halfway between two successiv~
zero passages of the burst signal. However, it is
assumed that owing to the inaccuracy of the horizon-
tal synchronizing pulse this trailing edge appears
at an instal~t t1. This trailing edge of the pulse
T triggers the multivibrator 52, so that said mul
tivibrator chan~es over (~ 5c). When at the in-
stant t2 a pulse Q from the zero passage detector
i9 applied to the gate circuit 41 a pulse (Fig. 5d)
is produced at the output of the monostable multi-
vibrator 42, which pulse is applied to the reset
,~ input 56 of the multivibrator 52. Th~s pulse resets
the multivibrator 52 to its original state at the
instant t2, so that the output signal of this mul-
.~ tivibrator.is the puise W shown in Fig. 5c.
.~ .
'' :
'
- . - 22
~ ' ' ' ' ~
. PHN 8201C
27.9.76
From the Figure it is evident that the pulse
width t o~ this pulse W corresponds to the time in-
terval t2-t1, so that it is a measure o~ the po~i-tion
of the trailing edge of the pulse T between tho two
c~ns0cutive zero passages of the burs-t signal. T.his
data is utilized by the invention or automatically
correcting the positi~on of the trailing edge o~ the
pulse T. The pulse width t~ of -the pulse W is measur-
ed with the aid of a measuring circuit 53 and convert-
ed into a correction signal ~or the monostable multi-
vibrator 51. In the example of Fig. 5 this correction
signal will result in a reduction of the pulse wid-~h
; Or the pulse T, in such a way that the trailing edge
., . of this pulse T is moved towards thQ instant to.
. ~ 15 (Fig. 5e), which in its turn results in a pulse W
from the multivibrator 52 in accordance with Fig.
~f.
Thus it is achieved that the leading edge
- of the keying signal W is always controlled towards
. 20 a position halfway be-tween two consecutive zero
passages of the burst signal, so that symmetrically
about this position a maximum tolerance is permis-
sible without giving rise to an erroneous time error-
. I ,
,I mea~urement. The multivibrator 52 preferably tak0s
the form of a monostable which in the absence of a
reset signal at the input 56 is automatically reset
. ' '
' . ' . :
~ 3
P~IN 8201C
` 27.9.76
:~'7~ 3 .'
after for example 140 nanoseconds
A practical embodiment of the monostable
multivibrator 5I in conjunction with the measuring
circuit 53 is shown in Fig. 6~ The measuring circuit
53 comprises a capacitance C2(4.7/u~) which is con-
nected to a voltage source V2 and which is charged
via a resistor Rl~(220 kohms). The charge of this
capacitance C2, i.e. the voltage acrpss it, depends ?
: on the signal on the base of a transistor T1. The ~,
base of the transistor T1 receives the output pulse
W of the multlvibrator 52 and this transis.tor con-
ducts for the duration o~ this pulse, so that the
voltage across the capacitance C2 Is determined by
`. the pulse duration of this output pulse W of the
: ~ 15 multivibrator 52. The voltage across the capacltance
: ~ . C2 is applied to an emitter~follower T2 and with the
aid of a resistor R3 (150 kohms) it is converted in-
to a control current which is appliod to the control
.terminal 57 of the monostable multivibrator 51.
This monostable multivibrator 51 comprises
an integrated multivibrator circùit 58 of the type
SN 741e3 (Sign0tics), whose time constant, i.e.
pulse duration, is determined by the capacitance
, C1(3300 pF) and the resistors R1(2.2 kohms) and
: 25 R2(6.2kohms). As the control input 57 is connected
: ~ to this capaci~ance C2 the charge on this capacitance- -.
. ~ .
C2 and thus the time constant depend on the correction
'
,
.
2l~ _ .
PHN ~201C
27.9.76
~r7Z~73
current. By dimensioning the resistors and capacitan-
ces as specified it is achieved that the pulse dura-
tion of the output pulse of this rnonostable multlvi-
brator circuit 58 as a ~m~tion o~ the correction
current can ~ar~ minimum appro~ 140 Danoseconds,
i.e. ma~imum hal~ the period of the NTSC colour
.
burst signal. It is obvious that many modifications
to this circuit arrangement are possible. For exam-
ple, it is also possible to use a variable delay
means instead of -the monostable multivibrator.
Fig. 7 shows a variant of the circuit ar-
rangement of Fig. 4, corresponding elements being
denoted by the same re~erence numerals. The output
signal of the limiter circuit 33 in this circuit
arrangement is not applied directly to the ero
passage detector 34, but is applied both directly
to an input 64~and via an inverting amplifier 61 to
an input 63 of a two-position switch 6~ whose master
contact 65 is connected to the zero passage detector
34~. This switch 62 i5 controlled by a control cir-
cuit 66 whlch receives the horizontal synchronizing
pulse train supplied by the sync separator 37 at an
:, :
input 67. This control circuit 66 then supplies a
symmetrical control signal of half the line fre-
quency to the switch 62 so that this switch is
changed over from line to line. Thus~ it is ensured
that the colour burst signals ~ consecutive lines,
- 25
PHN8201C
27.9.7G
~ 6~3
which are applied to the zero passage detector 34'
always have the same phase with respect to each other,
This means that this zero passage detector 34l no
longer need be of the absolu-te type, as in tho ver
sion in accordance with Fig. 4, but should merely
respond to the zero passages which occur in the
case of one specific sign of -the slope of the colour
burst signàl. This moans that this zero passage detec~
tor 34' may then taka the form of a simple monostable
multivibrator.
As pr0viously stated, the in-vention is by
no means limited to optical read apparatus, but is
also applicable.to read apparatus for a magnetic re-
cord carrier in the form of a tape. ~Irthermore~ the
circuit arran~ement of the time error.correctivn sys-
tem is not limited:to the system described, In prin-
cipl0, any suitable time error correc-tion system,
whether electromechanical or electronic, may be used,
the ultimate choice being generally determined by the
., ;
type of read system.
Furthermore, the in~en-tion is not limited to
- the coding system des~ribed by way of example, in i
which a complete standard NTSC colour television sig-
,
nal is frequency modulated on a carrier wave. The
invention may for example equally be used with cod- ;
îng systems in which, during recording, the c~romi- ~
:
~ nance signal is extracted and transposed -to a lower
:
: frequency band~ i.e. is record.0d as a modulation of
- 26 -
.~ , - ' ' .
.
PHN 8201C
27~9.76
.
a separate chro~ninance carrier wa~e, whilst the lumi-
nance signal lS frequcncy modulated on a carrier wave
of comparativel.y high frequency. In such coding 5yg-
tems the ~requency of the s~pa~ate chrominance carrier
is generally arl integral multiple of half the line
f`requency. During reproduction the read-out chromi-
nance signal is than re-transposed to the standard
frequency band by mixing it with a suitable mixing
frequency, so that again a standard NTSC or PA~
colour television signal is obtained. By ensuring
that the mixing frequency has the same time errors
as the read-out chrominance signal, the influence
of these timing errors on the ultimately obtained
standard colour television signa~l is substantially
: 15 reduced. By the indicated choioe of the frequency of
the separate chrominance carrier it is achieved that
.,
as pilot tone for producing this ~ixing frequency
use can slmply be made of the read-out horiz.ontal
synchronizing pulse train as is for example compre-
hensively described in United States Patent Speci.fi-
cation 3,803,3l~7 (PHN 4978). Instead of this horizon-
. ~ .
.~ ~; tal synchronizing pulse train it is then obvious
.
that use can also be made of a æero passage of the
colour burst signal with the aid of ~ device in ac-
J
cordance with the present invention. In that case it
; ~ is also possible to~use.the colour burst signal pre-
~ ' ~
..
- 27
.
PHN 8201C
27 . 9 , 7G
'Z6~;~
sent in the colour signal, or an additional burst si.g-
nal on the backporch of -the ho:rizonta:L syncllronizing
pulses, which has been added -to the luminance signal
during recording.
Furthermore it is not necessary to separate
the pilot tone from the;...complete decoded television
signal (output of decoder 18). Dependent on the cod-
ing system this pilot tone could be -sepa~ateds~1ome-
where wi'thin the decoding process.
It is to be noted that although the invention
will primarily be used to advantage ~or.reading out a
colour televislon signal, this invention may also be
~ used for reading out a black-and-white television
- signal. For this purpose, it will only be necessary
to add burst signals on tha backporch of the horizon-
tal s~nohronizing pulses during recording.
Fina].ly, the possibility will be described
of adding an additional burst signal to the televi~
sion signal during recording, which burst signal
may be utiliz0d for accurately measuring the time
: errors during r~production, This possibility is
represented in Fig. 8. This Flgure again sho~ys a
horizontal synchronizing pulse S of a television
slgDal with a horizontal blanking level:VB and a
peak level VT. However, prior to recordingl a
burst signal E has been superimposed on this hori~
zontal synchronizlng pulse S, i.e. on the peak level
-. ' ~ .' ,
~ 28
'.~ ' . '
' '
PHN 8201C
27.9.7G
~ 7~
VT thereof. This burst signal L has a frequency which
is an integral multiple of half the line frequency
and preferably an integral~mllltiple of the whole line
frequency.
During reproduction of the recorded tele-
vision signal this additional burst signal is ex-tract~
ed, after which in a similar way as described herein-
before with respect to the NTSC colour burst signal
the positionof a specific zero passage of this addi~
tional burst signal can be detected during each line
period and can serve for measuring time errors. ,~
The use of this additional burst signal is
of special importance when recording and reproducing
a colour television signal which complies with the
PAL standard. In this case the colour burst signal
cannot simply be used in the described time error
measuring system because the frequency ~f said PAL
colour burst signal lS an odd multiple of 1/l~ fH and
moreover exhibits a 25-Hz offset.
If furthermore an additional burst signal
with a frequency equal to a multiple of the line
frequency is used, a~non-absolute zero passage detec-
tor may be used for detecting the zero passages, i.e.
a detector which detects only the zero passages which
correspond either the positive-going or the negative-
going edges. Thus, the time interval within which the
beginning of the keying signal should occur is
- ~ .
~ 29
PIIN 8201C
27.9.76
~ 3
doubled compared with the use of the NTSC colour burst
signal, because this now corresponds to a f`ull period
of' the burst signal. Finally, the use Or this additional
burst signal superimposed on the horizontal synchroniz-
ing pulse has the advantage -that -the first -time i.nter-
val, i.e. the time between the leading edge of' the
horizontal synchronizing pulse and the koying signal
is substantially shorter than when the colour burst
signal is ùsed, This shorter time interval can be
realized more accurately with the aid o~ multivibrator .
cir~uits, which adds to ths reliabiLity of ths systsm.
,~
.
. ..
~, .
'
1~ .,
. ~_ 30 _ .
