Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
~L276~9~)
Background of the Invention
~ he present invention relates to a method for
producing screen printing blocks of the kind having
~creens o~ optional ~creen angle and screen line spacing
by line-wise optoelectronic scanning of an original for
producing an image signal and by line-wise recording by
means of a recording element displaced over and with
respect to a recording medium~ the latter having associa-
ted with it an orthogonal co-ordinate system subdivided
into areal elements and aligned in the line direction,
the locus co-ordiantes of the areal elements traversed
momentarily by the recording element being determined
continuously and a recording signal being generated for
the recording element by current comparison of the image
signal with a screen threshold signal, the recording
signal controlling the recording of the individual screen
dots as a configuration of areal element~ in the co-ordinate
system, and to a system for carrying out the method.
l`he inventive method is applicable for e~ample
in the case of a co~lour scanner for producing c~rrected
2Q colour separations. ~In the case of a colour scanner of
this nature, ~hich is known per se, a coloured original
is scanned point by point and line by line by means of an
optoelectronic scanning element and three primary colour
signals are concomitantl~ obtained which are converted
:
-- 2 --
in a colour computer into the colour-corrected colour
composition signal~ for recording the colour components
"magenta", "cyan" and "yellow".
Recording elements in the form of light ~ourcee
modulated in brightness by the colour extract ~ignal~,
perform the point-by-point and line-by-line e~po~ure o~
the colour component~ on a pho-tosensitive recording me-
dium. The colour compositions may be produced as half-
tone colour compositions for further processing i~ en~
graving machine~ or else as screen colour printing blocks
if they are t,o be applied as formes for colour of~set
printing.
The printing in superimposition of the differently
inked screen printing blocks of a colour composition ~or
multicolour reproduction is then performed in a printing
machine.
A moir~e pattern is generated since it is impossible
in practice to print the screen dots of the individual
component colours precisely on each other. A moiré
pattern of this nature makes itself felt in disturbing
manner in particular upon inspecting the finished printed
picture.
The obtrusiveness of moiré effects is reduced
in known manner, by the fact that the screen grids of the
individual colour components of a colour composition are
printed in superimposition in angularly staggered position
with respect to each other. By virtue of the screen angle,
the moiré phases formed are in effect either too small or
too large to be noticed as troublesome by the human eye.
Colour compositions wherein the individual screen grids
are turned through different screen angles with respect
to the recording direction, are required for a screen
rotation of this kind.
Consequently, four different screen angles are
needed for the four colour monochromes. To produce a
~27~90
.
-- 3 --
moiré minimum, it proved to be advantageous in four-
colour printing to select the screen angle -15 for
"magenta", the screen angle +15 for "cyan", the screen
angle 0 for "yellow" and -the screen angle ~45 for
"black". The screen angles ~hould be adhered to ver~
preci~ely since troublesome moire e~ects already in-
tervene at small angular deflections.
Other screen angles are then required complementarily
if other colour~ are to be printed, other print carries
are to be applied or if different screen line ~pacings
are to be printed one over another.
The direct application of a screen on half-tone
originals in the colour ~canner may for e~ample be per-
formed b~ means of a so-called contact screen application,
wherein the recording beam is complementarily modulated
by the density variation of a contact screen film posi-
tioned between the recording element and the recording
facility, to generate the screen grid elements.
For example United States Patent Specification
~o. 3,688,0~ discIosed a method for so-called "electronic
screen application" wherein each screen grid element is
built up in the manner of a picture pattern from indi-
vidual picture elements or type or body lines. The pic-
ture patterns of the different screen grid element sizes
are stored as recording data for all tonal values and
for different screen angles. ~he recording data are
currently read out and recorded in each case, which
correspond to the tonal values determined during scanning
of the original, during the reproducing operation.
Whereas the instrument-related composition screen
grid in which the screen grid elements are recorded is
aligned orthogonally in the recording direction and feed
direction of the implement, printing screen grids turned
in different degree with respect to the composition
screen grid are desisive for the precise positional
~Z7~90
4
location of the screen grid elements on the recording
medi~.
What is required is to fit the different printing
screen grids into the system of the printing lines. ~his
is particularly lmcomplicated according to United States
Patent Specification No. ~,657,472 if the tangent o~ the
screen angle is a simple rational number~ A common
areal element which ha~ the ~undamental ~truc-ture o~ the
screen pattern and which i9 repeated periodically on the
recording medium in the recording and feed direction,
whereby the recording operation i9 controllable by meane
of uncomplicated cadencing systems which are coupled to
the displacement o~ the recording medium or wlth the feed
motion of the recording element, then results for both
t5 ~creen systems in the case of such "rational ~creens".
Screen grids having screen angles whereof the
tangent is irrational, cannot be recorded according to
the method hereinabove described, so that the screen
angles of plus and minus 15 required for a moire minimum
cannot be established either.
A different method9 whereby "irrational ~creens"
may also be recorded9 is described in the United States
Patent Specification No. 3,922,484. In this known method,
XY pulse serie~ are derived ~rom the di~placement o~ the
recording drum and from the ~eed motion of the recording
element, the analysis of said series determining the
momentary positional locus of the recording element with
respect to the recording facility in an orthogonal co-
ordinate system aligned in the recording and feed direction.
~0 The XY pulse serie~ are converted in accordance
with a predetermined function, to generate a screen sig-
nal. This ~unction, which is periodic and bi-dimensional,
represent~ the screen pattern turned through the required
screen angle.
~5 During the recording action, the screen signal
. .
~7
-- 5 --
and the image signal are compared continuously and the
decision as to whether a screen ~rid element i~ to be
or is not to be recorded at the locus characteri~ed
by XY pulse serie~ i~ derived from the comparison.
The func-tion i9 reproduced electrically in
a function generator wherein, among others, other pul~e
series are initially generated by multipllcation o~ the
frequencie~ of the XY pulse serie~ by particular *actor~,
the factors being irra-tional or almost irrational and
representing different function~ of the screen angle
selected for the printing operation.
The multiplication is performed by means of phase-
locked loop circuitæ which, according to experience, have
a build up action and relatively low stability. The
required screen angle may consequently be adhered to with
a limited precision only, 90 that as already stated,
troublesome moiré phenomena may appear at a particular
angular deviation.
To improve the definition and printability of the
screen dots, it i9 frequently desirable to produce di-
fferent screen dot shapes or to split the screen dot into
partial elements, in accordance ~th United States Patent
Specification NoO 3,922,484 hereinabove referred to.
In the method disclosed by this last mentioned
Patent specification, it may well be possible to produce
circular or rectangular screen dots by mean~ of diff~rent
functions, but the possibilities of variation are ve~y
limited. Furthermore, some of the functions specified
may be reproduced in a function generator with dif~iculty
only, which is considered to be disadvantageous.
In the known device, the recording is produced
by several partial beams situated one beside another,
which are emitted from a recording element. The image
~ignal must be compared to different screen signals,to
control the partial beams. The generation of the screen
_ ... .. . . _ . . . .. .. .
o9~
signals which must make allowance for the different
points o~ impingement of the partial beams on the re-
cording facility, is not described in particular.
Summary of the Invention
Aceordlngly, in a method of the kind here~nabove
specified, the invention con6is-ts in that an or~hogonal
X-Y co-ordinate system which includes the screen angle
with the U~V co-ordinate system aligned in the line
direction, and i8 aligned in the direction of the screen,
i8 associated with the turned screen which is to be
recorded, said screen consists of orthogonal screen grid
element~ corresponding in size to the predetermined screen
line spacing, and each screen grid element consisting of
the areal elements with which are associated corresponding
~;y locus co-ordinates, and notwithstanding the screen
angle ~, a screen threshold value is associated in each
case with the areal elements of at least one spurious
screen grid element of optional screen line spacing as
a function of their x,y locus co-ordinates, said u;v
locus co-ordinates of said areal elements alocated during
current co-ordinate determination to a screen grid element
~hich is to be recorded and has a predetermined screan
line spacing, are recalculated into the limited range
of values of the corresponding ~;y co-ordinates o~ said
spurious screen grid element, and said screen threshold
value associated with each pair of co-ordinates which
upon comparison with the corresponding lmage signal de
termines whether the areal element in question is or is
not recorded as a part of a screen dot in the U-V co-
ordinate system, is determined by means of said recalculated
~0 or conYerted ~;y locus co-ordinates.
~ his method avoids or minimises the disadvantages
referred to.
- One advantage of the method specified consists in
that any optional screen angle, i.e. a screen angle whose
-. .. ~ . . : -
.
-
.
~L:12~
-- 7 ~
tangent is rational or irrational, may be set up with
high precision. It is thus possible to record "rational
screens" and "irrational screens". The screen angles
of plus or minus 15 and plu9 or minus 45 may pre~eren-
tially also be set up, a~ requir~d ~or a moiré mlnimum.The screen line spacing is concomita~tly una~fect~d by
the ~creen angle selected.
Brief Description of the Drawin~s
In ordor that the invention may be more clearly
understood, reference will now be made to the accompanying
drawings which show certain embodiments thereof by way
of example and in which:-
Fig. 1 shows a fundamental block diagram of acolour scanner,
Fig. 2 shows an enlarged section of the recording
medium,
Fig. 3 shows an embodiment of converter stage,
Fig. 4 shows another embodiment of converter stage,
Fig. 5 shows an embodiment of recording element,
Fig. 6 shows an advan~ageous development of the
converter stage,
Fig. 7 shows an embodiment for a psedo-random
cadence generator,
Fig. 8 shows a modified form of a colour scanner~
and
Fig. 9 shows an embodiment of random cadence genera$or.
Detailed Description of Preferred Embodiments
Referring now to the drawings, Fig. 1 shows a
fundamental block diagram of a colour scanner for the
production of electronically ~creened and correct colour
compositions.
A scanning drum 1 and a recording drum 2 are coupled
via a spindle 3 and are driven jointly by a motor 4 in
the direction of an arrow 5.
A colour original 6 i9 clipped on the scanning drum
1 and is scanned by a point of light of a light source
.. ____ _._ .. , .. . . . ,~ ..
- : .
' . .
:: :
3 lZ71~9~
-- 8 --
not shown in particular, point-by-point and line-by-line.
In the case of an opaque original it is the reflected
scanning beam, and in the case of a transparent original
it i9 the transmitted scanning beam, which reaches a
scanning element n , being modulated in brigh~ness by the
pictorial content of the original 6. The colour ~ignal~
R,G and B which represent the colour components of the
image dots scanned, are generated in the scanning element
7 by colour separation by means of colour filters and of
optoelectronic conversion o~ the scanning beam.
The scanning element 7 is di~placed parallel to
the scanning drum 1 in the direction of an arrow 10~ by
means of a motor 8 and a spindle 9.
The colour analogue signals, R, G, B pass ~rom
the scanning element 7 and via a post-connected amplifier
11 to scanning printing transducers 12, 13, 14 in which
they are converted by means of a cyclic scanning sequence
TA into digital colour signals R', G' and B' having a
word length of 8 bits for example, a scanned picture dot
being co ordinated with each cycle of the cyclic scanning
sequence TA.
~ he cyclic scanning sequence TA is generated by
frequency division in a divider stage 15 from a cyclic
sequence ~0, which is generated by means of a cadence
generator 16 coupled in rotation to the drums. The cyclic
scanning sequence i9 fed to the A/D transducers 12, 13
and 14 via a conductor 17.
The digital colour signals R', G' and B' are
converted in a digital corrector circuit 18 into the correct
monochrome signals Mg, Cy, Ye for recording the mono-
chomes "magenta", "cyan" and "yellow".
A colour and/or graduation correction is performed
in the digital corrector circuit 18, depending on the
- requirements of the reproduction process. A corrector
~5 circuit o~ this nature is described exhaustively, for
L2~9~
example in the United States Patent Sp~cification No.
3,885,244.
A digital memory for intermediate storage of -the
monochrome signals may also be post-connected to the
corrector circuit 18, to perform a scalar change between
the original and the rscording in accordance with
United States Patent ~pecification No, 3~272~g18~ or
to record -the pictorial content of the entire original
and to recall or repeat the same for recording with a
time lag or if appropriate at a specific locus.
In the embodiment, the digital monochrome signals
Mg, Cy, Ye~ reach a colour separation switch 19 whereby
one of the digital monochrome signals is selected in
each case for screened recording of a chromatic selection.
The invention is obviously also applicable if al1
the chromatic 6elections are recorded in one operation,
beside each other in parallel or serially, on the cir-
cumference of the recording drum 2.
A recording element 20 is displaced by means of
another motor 21 and o~ a spindle 22, axially along the
revolving recording drum 2 in the direction of the arrow
10, The recording element 20 perform the point-by-point
and line-by-line illumination o~ the scrsen dots on a
photosensitive recording medium 23 which i9 arranged on
the recording drum 2.
The recording beams 24 focussed on the recording
medium 23 by the recording element 20 produce a number of
e~posure points Pn which by virtue of the relative dis-
placement between the!recording element 20 and the re-
cording drum 2 also illuminate the recording medium 23along type lines 25 extending in the circumferential
direction (recording direction).
Each screen dot 26 comprises a number o~ such
closely set type lines 25. The size and shape o~ a screen
dot depends on the length of the type pr body lines 25
_ _ _
~ L;271~
1 o --
or rather on the momentary period o~ energisation of
the individual recording beams 24~ The recording
beams 24 may be switched on and of~ by recording sig-
nal~ A which are fed to the recording element 20 via
conductor3 27. One embodiment of recording element 20
i9 illustra-ted in Fig. 5.
It i~ within the scope of the in-ve~ltion to
illuminate the type line~ 25 of -the screen dot 26 by
means of a single recording beam 24 de~lectible trans-
versely to -the reoording direction.
In this case, the screen dot 26 comprises type
lines e~tending transversely to the recording dlrection.
The deflection of the recording beam 24 may be produced
by means of an electro-acoustic de~lector system, e.g.
as specified in the United State~ Patent Specification
No. 3,725,574.
The process stages for obtaining the recording
~ignals An are to be described in particular in the
following.
The momentary positional locus of the e~posure
points Pn on the recording medium 23 is established on
the recording drum 2 by means of an instrument-related
U-V co-ordinate sy~tem 28 unaffected by the screen angle
beta, whose U axis is aligned in the peripheral direction
of the recording drum 2 and whose V axis is aligned in
the ~eed direction of the scanning and recording elements.
The U-V co-ordinate system 28 is subdivided into a
plurality of areal elements from which the ~creen dots
which are to be recorded are built up.
The positlonal locus of the screen dots 26 on the
recording medium 23 is given by a screen grid 29 in an
X-Y co-ordinate system 30 which is turned through the screen
angle ~with respect -to the U-Y co-ordinate system 28.
The screen grid 29 comprises a plur~lity o~ screen
grid elements whose size depends on the screen line ~pacing
. :'
:
Z7(;~
- 11 _
which is to be recorded. ~ach screen grid element i8
built up from the areal ele~ents which are associated
with corresponding x';y' locus co-ordinates.
A spatial function R = g(x;y) having a range o~
value~ limited to the 3purious screen grid element,
which defines the size of the s¢reen dots as a ~unc-t~on
of different image signal amplitudes ~tonal value ~tage~)
and the screen dot shape, is preset for a ~purious screen
grid element which is una~ected by the screen angle and
the screen line spacing of the screen which is to be
recorded. In this function, R is the screen threshold
value of an areal element, and x;y are its associated lo-
cus co-ordinates in -the X-Y co-ordinate system 30.
The range of values of the x-y locus co-ordinates
appertaining to the preset function is restricted as
compared to the range of value~ of the x';y' locus co-
ordinates of the exposure points P~l determined upon tra
versal of the entire recording area.
The spatial representation of the function
R = g(x;y) is also re~erred to as a "screen hill" whose
base surface fills the spurious screen grid element and
wherein a cross-sectional surface passing through the
screen hill at the level o~ the momentary image signal
amplitude indicates the screen dot size for the tonal
value in question.
In the course of reproduction, the current ~
locus co-ordinates of the exposure points in the X-Y
co-ordinate system 30 are determined, converted to the
limited range of values of the x;y locus co-ordinates o~
the spurious screen grid element, and the screen threshold
value co-ordinated by means of the function is called up
or invoked~ The screen threshold value is compared to
the image signal and the decision whether the areal
- element in question in the U- Vco-ordinate system 28 is or
is not to be recorded as part o~ a screen dot, is derived
-
,
. ~.. 1~27~
- 12 -
from the comparison.
The U and the V axes are divided into fundamental
step~ Q u and ~ v , -to determine the locus co-ordinates
un;vn of the exposure point~ Pn in the U V co-ordinate
~ystem 28. The length of the fundamental step~ may
differ between the axes.
The locus co-ordinates un;vn amount to multiple of
the fundamental steps~ u and ~v.
In a fir~t process stage, the momentary locus
co-ordinate~ un;vn of the exposure points Pn are deter-
mined by current counting or summating addition of the
fundamental steps ~u and ~b by means of two timing se-
quence Tu and Tv in a converter stage 31. The timing
sequence Tu is obtained by frequency division in a divi-
der stage 32 from the timing sequence To of the cadencegenerator 16 and is fed to the converter stage ~1 ~ia a
conductor 3~. A fundamental step~ u is co-ordinated with
each cycle of the timing sequence Tu. The length of
the fundamental step may be changed by the frequency of
the timing sequence Tu and may if appropriate be adapted
to the required precision.
A circumferential pulse emitter 34 which is
equally coupled to the recording drum 2, generates a
circumferential pulse Tv co-ordinated in each in3tance
with a fundamental step~v, once per revolution, i.e.
after every feed step of the scanning element 7 and of the
recording element 20. The circumferential pulses T are
fed to the converter stage 31 via a conductor 35.
The locus co-ordinates u1~v1 for the first point
of exposure P1 are derived from the equation:
u1 = Cu ~u ( 1 )
V1 - CV ~V
~ u andL~v denoting the fundamental steps in the
U-V co-ordinate system 28 and Cu and Cv denoting the
.. . . ... . ..
Z7~39
_ 1 3
number of timing pulses Tu and ~v re~pectively.
~ he pairs of locus co-ordinates for the other
e~posure points may advantageously be calculated from
the pair of locus-co-ord~nates of one of the exposure
point~, e.g, of the ~irst exposure poin-t P1. 'rhe
position of the e~po~uxe points Pn wi-th respect to
e.ach other may be optional, but the expo~ure points will
commonly lie on a ~traight line.
To es-tabli~h an homogenou~ density curve over
-the screen dot surface, the ~traight line corre~ponding
to German Patent Application No. P 26 53 539.7 extenda
at an angle to the generatri~ of the recording drum 2.
In this case, the mutual spacings u* and v* o~ the
e~posure points are constant and depend only on the stru-
ctural design of the recording element 20 and on the
scale of reproduction. The locus co-ordinates un;vn
of the other e~posure points P~ may consequently be
calculated in accordance with the equation un=u1+(n-1)u*
and vn v1~(n 1)v -
The e~posure points are frequently situated on
the actual generatrix of the recording drum 2, however
when u* = 0.
Since the function R=(x;y) is preset notwithstanding
the screen angle ~ , and the screen line spacing, the
locus co-ordinates un;vn of the U-V co-ordinate system
are currently converted into the corresponding locus
co-ordinates ~'n; Y'n of the X-Y co-ordinate system 30
in a second process stage, in the converter s-tage ~1,
with allowance for the screen angle ~ and for the
different screen line spacings of the screen grid element
which is to be recorded and of the spurious ~creen line
element~
. ~.
During the conversion, the greater range of
values of the locus co-ordinates xn;yn arising under
~5 illumination of the entire surface of the recording medium
_ . . .
.
23, is simul-taneously restrlcted to the limited range
of values o~ the 2; y locus co-ordinates of the preset
function R = g(x;y) . Thi~ operation will be de~cribed
in particular in the following.
The conversion o~ the locus co-ordinatea in the
converter ~ta~e 31 is performed in accordance with the
equations :
x n Ku un- C09 ~ ~ KV.Vn.sin ~- M
(2)
Y n KU-Un-Sin~ ~ Kv.vn.cos~ _ M
In the equations (2), the coefficients Ku make allowance
for the different screen line spacings of the screen
grid element which i~ to be recorded and of the spurious
screen grid element, and the terms M~ and My take into
account the limitation of the current ~';y' locu~ co-
ordinate~ to the value range of the function.
The screen angle and the coe~ficients are pre~etat the programming input terminals 36 and 36' of the
converter stage 31.
E~amples of embodiment~ of converter stage 31
are depicted in Figures 3 and 4~
At its output terminals, the converter stage
31 determines corresponding pa;rs of co-ordinates xn;yn
for each exposure point Pn. From the pairs of co-
ordinates xn;yn, and in accordance with the preset fun-
ction R-g(x;y), the screen generators 37, 38 and 39
generate corresponding digital scree1l threshold values
Rn which9 like the digital chromatic composition signals 9
equally have a word length of 8 bits.
Digital comparators 42, 4~ and 44 are incorporated
~or comparing the screen threshold value3 Rn on the
- conductors 40 to the colour composition ~ignal ~elected
on the chromatic selection switch 19 on a conductor 41.
These comparators 42; 43 and 44 generate the
recording signals An on the conductors 27~ with which
the illumination of the screen dots 26 on the recording
.. .. . . . .. . . .. . ~ .......
.~.Z7~g [/
- 15 -
medium 2~ i~ controlled.
A variety of advantageous po~sibilities i~
available for the etructure o~ the 6creen generatora 37;
38;39.
In the embodimen-t, the ~creen generator~ comprise
read-only storage units, in which the same ~unction
R = g(x;y) is stored in each case.
The read-only storage unit comprise~ a storage
matrix, e.g. incorporating 32~32 storage cells for the
screen threshold value3. The ~torage cells are selectible
by means o~ 32 ~ addresses (5 ~t ) and of ~2 ~ addre3~es.
In this case, the x;y value range for the function is
limited to "32", i.e. to the addresses 0 to 31 in each
case.
It may also be envisaged to address all the read-
only memories with the ~;y co-ordinate values of one of
the e~posure points and to obtain the different screen
threshold values R for ths other exposure points by making
allowance for the appropriate mutual spacings u* and v* of
the other e~posure points converted into the X-Y co-ordinate
system 30, when programming the individual read-only
memories.
To save on read-only memories, the differ~nt
pairs of x3y locus co-ordinates for the exposure points
may addres~ a single read-only memory consecutively by
the time~sharing method.
The screen generators 37, 38 and 39 may equally
comprise function generators which reproduce the function
R = g(~;Y).
In this case, the function could preferentially
assume the form R=g(D~x~E.y).
In the case in which the function generator
operates digitally, the function R=g~;y) could be stored
in a memory whose address input terminals have applied
to them the sum (D.~ ~ E.y) . Identically, the products
- ~ ,
- ': .
~z~v9~
- 16 -
(D~) and (E.y) may be stored in one or more memories,
which may then be addressed direct with the x;y co-ordinate
values.
In the arrangement according to Fig. 1, the
feed displacement of the scanning elemen-t 7 and recording
element 20 in the direction of the arrow 10, may be
intermi-ttent or continuous.
In the case of an intermittent ~eed J the scanning
and recording actions occur around the drums along cir-
cular image lines whose mutual ~pacing corresponds to afeed step. By contrast, in case of a continuous Peed,
the scanning and recording actions occur along image lines
extending helically around the drums. In this case, small
errors arise during the recording operation, which in
accordance with an advantageous development of the in-
vention principle may be cancelled in the conversion
equations (2) by correction factors (Sv.sin ~) and
(SV-c09 ~ tSvll denoting the pitch o~ the helix and 11~ "
- again denoting the screen angle. The conversion equations
then have the following form:
x=Ku.u.(cos~ + Sv.sin ~) ~ K~.v.sin ~ M~
y=Ku.u.(-sin ~+ Sv.cos~ ) + Kv.v.cos ~ My (3)
For a clearer grasp of the screen grid element
recording, Fig. 2 shows an enlarged section of the re-
cording medium 23 with the instrument~related U-V co-
ordinate system 28 (U direction = recording direction)
and with turned screen grid 29 which is to be recorded
and with respect to this the X-Y co-ordinate system 30
is aligned, the co-ordinate systems including the screen
angle ~.
The screen grid element 47 of the turned screen
grid 29, comprising the screen dot 26, to a degree re
presents the fundamental structure of the screen pattern
which is continued periodically in the X and Y directions
throughout the recording suxface.
.. _ . .. ... . . _ . . _ _ _ ... ~ _
7~0
- 17 -
The screen dot 26 compriseæ a number of mutually
ad~acent type lines 25 extending in the recordi~lg direc-
tion. Each type line 25 is built up from individual
areal element~ 48 which are a~sociated with current u;v
and x';y' locus co-ordinates.
A ~purious screen grid element 4g o~ optional
screen line ~pacing which equally compri~es a number of
areal elements 50, is also shown. ~ach areal element 50
is associated with a screen threshold value R and with
a pair of x;y 10CUB co-ordinates, whose range of values
is restricted however to the spurious screen grid element
49.
~ or each areal element 48 which is momentarily
traversed by an exposure point a 3creen threshold value
appertaining to a congruent areal element 50 in the
spurious screen grid element 49 is de-termined in accordance
with the equations (2) specified in Fig. 1, and is com-
pared to the image ~ignal to obtain the recording sig-
nals, during the recording operation.
Different possibilities arise for obtaining the
image signal.
In the embodiment according to Fig. 1, the
recording element 20 which is merely hinted at in ~ig. 2,
generates for example three recording beams 24 and thereby
also several mutually adjacent exposure points Pn which
simultaneously illuminate a corresponding number of type
lines 25 during one revolution of the recording drum 2.
If three exposure points P1 to P3 are present, as
shown in Fig~ 2, and if the screen dot 26 comprises six
~0 type lines 25 (or linear tracings), the screen dot 26
has been exposed by the scanning element 7 and recording
element 20 after two drum revolutions or rather feed
steps. In this case, no more than two image data of the
original 6 scanned on two mutually adJacent image lines
51 are available for all the linear tracings 25 of the
.
. ' .
~ ~Z7~
- 18 ~
screen dot 26. The precision of the recording may be
increased if an lmage datum obtained from a positionally
co-ordinated image line 51 is avallable ~or each linear
tracing 25.
This may be accomplished in advantageous manner,
in accordance with the United States Patent Speci~ication
No. 4,149,195, by the fact that a plurality of image
dots which are mutually adjacent in the V direction of
the UV co-ordinate system 28 are scanned simultaneously
in the original 6 and that it is the image signal of the
image dot whose positional locus on the original 6
corresponds to the linear tracing 25 which is just to
be recorded is in each case selected for controlling the
recording element.
The recording element 20 may however generate
no more than one recording beam 24 and thus at the same
~ time no more than one exposure point P1 on the recording
facility 23. In this case 9 one linear tracing 25 is
illuminated in each case per revolu~io~ of the recording
dru~ 2, the scanning element 7 and the recording element
20 performing one feed step by a linear tracing width
after each revolution. An image datum is thereby obtained
from an image line 51 of the original 6 which is posi-
tionally co-ordinated in the V direction, for each linear
tracing 25 of the screen dot 26~ This method may well
be very precise, but it operates very slowly.
It is obviously also possible to scan several image
dots in circumferential direction for each screen dot 26.
Fig. ~ shows an embodiment of the converter
stage 31 wherein the current u;v locus co-ordinates of
the U~V co-ordinate system 28 are determined by counting
the fundamental steps ~u and ~v and are converted in
- accordance with equation (2) into the co-ordinates xn;yn
for driving the screen generators 37, 38 and 39. ~~
~he values Ku.~ u and Kv. 4 v, as well as cos ~ and
7~9~
- 19 -
~in ~, are stored in storage registers 53 to 56.
The cycles Tu and Tv on the conductor~ 33 and
35 are counted in the counters 57 and 58. The counter
levels correspond to the factors Cu and ~v- The *a¢tor~
are multiplied in the multiplication stages 59-62l in
accordance with equations (2), and the products are
then summated in the adding sta~es 6~ and 64. The
current locus co-ordinate3 xl1;y'1 for the first exposure
point P1 are the result as a 32-bit datum.
Since the 32 x addresses and 32 ~ addresses o~
the read-only memory as selectible in the screen genera-
tors ~7, 38 and 39 by means of 5-bit data in each case,
the calculated locus co~ordlnates ~'1;Y'1 ~2-bit) are
converted to the limited x1;y1 address range from 0-31
(5-bit) according to the relationship ~ 1'mod. 32 or
resp.y1=y~1 mod-32, in the stages 65 and 66. The
conver~ion occurs by stripping or skimming off the bits
of higher value.
~he output signals x1 and Y1 of the stages 65 and
66 are the pair of addresses for the exposure point P1
for selection of the read-only memory 37.
The other pairs of addresses xn;yn for the other
exposure points Pn are determined by adding the values (n-1)x and
(n~1)y to the caIculated locus co-ordinates x~1 and Y'
in the adding stages 67 - 70, and by ~tripping o~f b1ts
in the ~tages 71 - 74. The values ~* and y* are calculated
from the predetermined spacings u* and v* of the exposure
points Pn.
The pairs of addresses xn;yn for the other
exposure points Pn may also be determined by adding the
values (n-1)u* and (n-1)v* to the locus co-ordinates
U1 and v1 of the first e~posure point P1 and by subsequent
conversion.
Fig. 4 shows anather example of embodiment of a
~5 converter stage 31 wherein the locus co-ordinates un;vn
27(~9~
- 20 -
o~ the exposure points Pn are determined by summatlng
addition of the fundamental steps u and v.
The values Ku.~ u.cos ~ , Ku.~ u.sin ~ , Kv.~ v.
sin~ and Kv. ~v.cos ~of equation (2) are stored in the
storage registers 75 -78.
For summating addition of these values, the
storage registers 75 - 78 are in each case connected to
the first input terminals of adding stages 79 - 82.
The adding ~tages 79 - 82 have post~connected to them
other storage registers ~3 - 86 whereof the output ter-
minals are in each case connected by return lines to the
second input terminals o~ the associated adding stages
79 - 82. The acceptance o~ the addition resul-ts in the
storage registers 83 - 86 is controlled by the timing
sequences Tu and Tv on the conductors 33 and 35.
The mode of operation of the adding stage 79
in combination with the storage register 83, is the
following. Assuming the contents of the storage register
83 to be nil, the addendum at the secondary input terminal
of the adding stage 79 is then also nil. The value Ku.
~ UCOS ~i8 consequently picked up in the storage register
83 with the first cycle of the timing sequence Tu on the
conductor 33~ This value i9 fed back to the secondary
input terminal of the adding stage 79 and added on
thereat, so that the value 2KU.~ u.cos~ is taken into the
storage register 83 with the second cycle of the timing
sequence ~u-
The content~ of the storage registers 83 and 84are added together in an adding stage 87, and those of
the storage registers 85 and 86 in another adding stage
88. The results are the locus co-ordinates x'l and Y'1
for the fir~t exposure point P1~ which are converted into
- the pa.ir of locus co-ordinates xt;y1 by stripping in the
stages 89 and 90.
As already described in respect o~ Fig. 3, determining
. _ . . . . .. . . .. . .
~Z709~
- 21 _
the pairs of locu~ co-ordinates xn;yn for the other
exposure points Pn i~ performed by means of the adding
stages 91 - 94 and by means of the stages 95 - 98.
~etermining the pairs of locus co-ordinate~ for
the other exposure points Pn may also be performed from
the known values u* and v* or el6e by appropriate
presetting of the s-torage register~ 8~ - 86.
Fig. 5 shows an embodiment of recording element 20,
A laser generator 101 generates a polarised light
beam 102 which consecuti-vely pas~es through three
partially transparent mirrors 103. The recording beam~
24 are deflected by reflection out of the beam 102 and
are directed at the recording medium 23 by adjustment of
the mirrors 103. A "twister" crystal 105, a polari~ing
filter 106 and an objective 107 are in each case situated
in the beam path of the recording beams 24. When the
twister crystals 105 are not energised, the polarisation
planes of the polarising filters 106 are turned through
precisely 90~ with respect to the polarising plane of
the recording beam 24, so that the latter is neutralised.
An electrical field i~ generated in a twi~ter crystal
105 by means of a control voltage between the control
electrode 108 and the counterele~trode 109, which is
at earth potential. The electrical field turns the
polarisation plane of the recording beam 24, in such
manner that the same no longer impinges on the subsequent
polarising filter under the blocking angle, so that the
recording beam 24 is activated.
The twister crystals 105 are thus utilised as
light switches which are activated and deactivated by the
digital recording signal -s An on the conductors 270
The recording signals An are converted via amplifiers
110 into the control voltages for the twister crystal~
1050
~5 Instead of the system of mirrors, a separate
~LZ70~0
- 22 -
laser generator 101 could also be present for each
recording beam 24. The recording beams 24 emerging
from the polarising filter~ 106 could als~ be focussed
on the recording medium 23 via optical ~ibre~0
In a modified embodimen-t, the recording element
20 may also consist of a line of light-emlbting diodc~,
each individual ~ED being controllable by means o~ a
recording signal An.
The method is applicable even if the screen dot~
are recorded on an appropriate radiation-sensitive medium
by means of a different source of radiation.
The ~creen generation may be improvea complementarily
by ~toring a greater number than 32 g 32 screen threshold
values in the read-only memories of the screen generators
37~ 38 and ~9. The improvement i~ advantageously accom-
plished even without a corresponding increase of the
storage capacity, if auxiliary values whose quantities
are determined in random manner, are superimposed on the
unconverted or converted locus co-ordinates of one of the
exposure points prior to addressing the read-only memories.
In the embodiment, these randomly selected
auxiliary values xh and Yh are added to the converted
current locus co-ordinates xl1 and Y'1 of the first ex-
posure point P1 according to the relationship
~ '1 + Xh
Y 1 Y 1 + Yh
Fig. 6 shows a pre~erred development of the converter
~tage according to Fig. 3, for application of this measure.
To simplify matters, only those functional groups which
contribute to an understanding have been taken over from
Fig. 3. The adding stages 63 and 64 are followed by
complementary adders 111 and 112 wherein the auxiliary
values ~h and Yh are added to the converted locus co-
ordinates ~'1 and Y'1 to obtain the new locus co ordinates
==_,
-
~ I.Z~
23
X~1 and y'1. The corresponding locus co-ordinate~ o~
the other e~posure points are then derived from these
locus co-ordinates. Such auxiliary values may also be
added to the calculated locus co-ordinates o-f the in-
dividual expo~ure points. ~he auxiliary value~ xh andYh a~e obtained in separate pseudo-random genera-tors 113
and ~14 and are fed to the corresponding adding appli-
ances via the output terminals 115 and 116. The input
terminals 117 and 118 of the p~eudo-random genera-tors
113 and 114 are timed by means of the timing sequence
Tu on -the conduc-tor 3~ (or by the timing ~equence Tv on
the conductor 35). Fig. 7 shows an embodiment o~ pseudo-
rando~ generator. The measures specified may evidently
also be taken in the converter stage according to Fig. 4.
Fig. 7 shows an embodiment for a pseudo-random
generator for generating the auxiliary values xh and Yh-
The pseudo-random generator 113;114 substantially
comprises an n-bit shift register 120 and a NOR feedback
circuit 212. The input terminals 117;118 of the shift
register 120 are acted upon by the timing sequences Tu
and Tv, respectively. ~epending on which of the output
terminals of the shift regist~r 120 are led back via
the feedback circuit 121, a quasi-random sequence of
output values which is repeated only within a considerable
period9 is generated at the output terminals 115;1160
A pseudo-random generator of this kind is
described exhausti~ely in the periodical "Electronics,"
of May 27th 1976, at page 107.
To improve screen generation, a timing sequence
T'u whose timing intervals are randomly generated, could
also be applied instead of a superimposition of auxiliary
values.
~ ig. 8 shows a modification of the system according
to ~ig. 1, in which a random cadence generator 119 is
situated bet~een the frequency divider 32 and the con~erter
`` 3L~Z'7
-- 2~ --
stage 31.
Fig. 9 show~ an embod:iment ~or a random cadence
generator 119. The timing ~equence Tu obtained in the
frequency divider 32 is fed to n time-lagging ~tage~ 122
with di~fering delay periods ~r
The time-lagging ~tages 122 are connected to the
input terminal~ 12~ of a multiplexer 12~ at whose output
terminal~ 125 the random timing sequence T'u io deli~ered.
A p3eudo-random generator 11~ or re~pectively 114 ac¢ording
to Fig. 7 i9 connected to the control input terminal
126 of the multiple~er 124.
