 |
Claims  |
|
|
We claim:
1. In imaging apparatus having first and second different imaging systems,
wherein said first and second imaging systems generate respective first
and second image depictions of a common original image, and said second
imaging system has a variable response, a method for obtaining an
appearance match between said first and second image depictions and for
generating operational settings for varying the response of said second
imaging system so as to calibrate a response of the second imaging system
to a response of the first imaging system, the method comprising the steps
of:
obtaining data values for corresponding portions of said first and second
image depictions in a pre-determined manner that substantially parallels
interpretative color preferences of a viewer; and
determining, in response to said data values and through a pre-defined
model of said second imaging system and pre-determined matching
principles, operational settings for setting the response of said second
imaging system to produce an image depiction, through said second imaging
system, that is an appearance match to said first image depiction, wherein
said matching principles comprise a plurality of pre-defined rules that
collectively define an appearance match between corresponding image
depictions produced by said first and second imaging systems in response
to a common source image applied as a common input to both said first and
second imaging systems.
2. The method of claim 1 wherein said data values obtaining step comprises
the steps of:
obtaining measurement data for the corresponding portions of said first and
second image depictions; and
transforming said measurement data into a pre-defined coordinate system
that represents color information in the predetermined manner to yield
said data values.
3. The method of claim 2 wherein said imaging system comprises either a
press or a first proofing system and said first image depiction comprises
a target image; and said second imaging system comprises a second proofing
system and said second image depiction comprises a proof image.
4. The method of claim 3 wherein said measurement data obtaining step
comprises the step of densitometrically measuring corresponding portions
of the proof image and the target image.
5. The method of claim 4 further comprising the steps of:
(a) obtaining, from associated portions of the target image, densitometric
data for: media density, three-color solid overprint densities, red and
green solid overprint densities, densities of a pre-defined halftone dot
size for three-color tints, and densities for a black solid and a black
tint;
(b) obtaining, from corresponding portions of the proof image,
densitometric data corresponding to all the densitometric data set forth
in step (a) above;
(c) generating, through first associated corresponding ones of the
densitometric data obtained through steps (a) and (b) above and in a first
pre-defined manner, recommended changes in solid area densities for each
of a plurality of process colors;
(d) generating, through second associated corresponding ones of the
densitometric data obtained through steps (a) and (b) above and in a
second pre-defined manner, recommended changes in halftone dot area size
for each of the process colors;
(e) generating, through third associated corresponding ones of the
densitometric data obtained through steps (a) and (b) above and in a third
pre-defined manner, recommended changes for solid and black tint,
respectively;
(f) generating the proof image, through the proofing system, using all the
recommended changes provided through steps (b)-(e) above; and
(g) repeating steps (b)-(f) above until an acceptable appearance match
results between the target image and a most recent proof image.
6. The method in claim 5 wherein said measurement data transforming step
comprises the step of converting each measurement triplet r-g-b (red,
green, blue) value from r-g-b coordinate space into a corresponding a-t-d
triplet value.
7. The method in claim 6 wherein said converting step utilizes the
following equation:
##EQU10##
8. The method in claim 5 wherein said pre-determined matching principles
comprise:
(a) for cyan, yellow, and magenta solid process colors:
first matching "a" coordinate associated with ones of said data values for
corresponding three-color solid overprint test patches in both the proof
image and the target image;
second matching, in a "t-d" plane, a hue angle associated with ones of said
data values for corresponding solid red test patches in both the proof
image and the target image;
third matching, in a "t-d" plane, a hue angle associated with ones of said
data values for corresponding solid green test patches in both the proof
image and the target image; and
(b) for three-color tint overprints:
fourth matching tint densities in either a-t-d or r-g-b coordinates at a
given dot size for data values associated with corresponding three-color
tint overprint test patches in both the proof image and the target image;
and
(c) for solid black and black tint;
fifth matching an "a" coordinate value in terms of "N-channel" density for
data values associated with corresponding solid black and black tint test
patches in both the proof image and the target image.
9. The method in claim 8 wherein said first matching step comprises the
step of matching the "a" coordinate in the a-t-d coordinate space
associated with said ones of said data values for the corresponding
three-color overprint test patches in both the proof image and the target
image according to the following equation:
N.sub.a.sup.T -(N.sub.a.sup.P +.DELTA.N.sub.a)=0
where:
N.sub.a.sup.P represents an "a" coordinate value for the near neutral
overprint in the proof image;
N.sub.a.sup.T represents an "a" coordinate value for the near neutral
overprint in the target image;
N.sub.a represents a resulting correction needed in a three-color overprint
"a" coordinate;
wherein said second and third matching steps comprise the steps of matching
the hue angle associated with ones of said data values for the
corresponding solid red and green test patches in both the proof image and
the target image according to the following respective equations:
R.sub.t.sup.T (R.sub.d.sup.P +.DELTA.R.sub.d)=R.sub.d.sup.T (R.sub.t.sup.P
+.DELTA.R.sub.t)
G.sub.t.sup.T (G.sub.d.sup.P +.DELTA.G.sub.d)=G.sub.d.sup.T (G.sub.t.sup.P
+.DELTA.G.sub.t)
where:
R.sub.t.sup.T and G.sub.t.sup.T represent "t" coordinate values for solid
red and green test patches for the target image;
R.sub.t.sup.P and G.sub.t.sup.P represent "t" coordinate values for solid
red and green test patches for the proof image;
R.sub.t and G.sub.t respectively represent a resulting correction needed in
the red and green "t" coordinate values;
R.sub.d.sup.T and G.sub.d.sup.T represent "d" coordinate values for solid
red and green test patches for the target image;
R.sub.d.sup.P and G.sub.d.sup.P represent "d" coordinate values for solid
red and green test patches for the proof image; and
R.sub.d and G.sub.d respectively represent a resulting correction needed in
the red and green "d coordinate values;
wherein said fourth matching step comprises the step of matching the tint
coordinates (N.sub.a, N.sub.t, N.sub.d) for data values associated with
the corresponding three-color overprint test patches in both the proof
image and the target image according to the following equations evaluated
at the given dot size:
N.sub.a.sup.T -(N.sub.a.sup.P +.DELTA.N.sub.a)=0
N.sub.t.sup.T -(N.sub.t.sup.P +.DELTA.N.sub.t))=0
N.sub.d.sup.T -(N.sub.d.sup.P +.DELTA.N.sub.d)=0
wherein said fifth matching step comprises the step of separately matching
the "a" coordinate value for data values associated with the corresponding
solid black and black tint test patches, the latter at the given dot size,
in both the proof image and the target image, both according to the
following equation:
K.sub.a.sup.T -(K.sub.a.sup.P +.DELTA.K.sub.a)=0
where:
K.sub.a represents "N-channel" solid or tint black density;
subscripts T or P representing the target image or proof image; and
K represents a resulting correction in solid or tint black density,
respectively.
10. The method in claim 9 wherein the given dot size is substantially 50%.
11. The method in claim 5 wherein the pre-defined model is a linearized
model of the response of said second proofing system at a given operating
point, said model comprising a plurality of sensitivity coefficients which
collectively define changes in expected response of said second proofing
system to pre-defined changes in said operating point.
12. The method in claim 11 wherein said changes in said operating point
comprise changes in solid area density values of cyan, yellow, magenta or
black process colors or halftone dot sizes associated with any of said
process colors.
13. The method in claim 12 wherein said pre-defined model is given by the
following equations:
(a) for three-color (red, green and blue) solid overprints:
##EQU11##
where: C.sub.r represents the change in cyan process color solid area
density for red light;
M.sub.g represents the change in magenta process color solid area density
for green light;
Y.sub.b represents the change in yellow process color solid area density
for blue light; and
N.sub.a represents a resulting change in an "a" coordinate value for a
three color overprint;
(b) for red and green solid colors:
##EQU12##
where: G and R represent resulting changes in green and red coordinates
(with subscripts "t" and "d" denoting the "t" and "d" coordinates,
respectively);
(c) for three-color tint overprints
##EQU13##
and (d) for black solids and tints:
.DELTA.K.sub.a =S.sub.n .DELTA.K.sub.n
where:
K represents a resulting change in an "a" coordinate value for black (K)
solid or black tint density; and
S represents a corresponding matrix of sensitivity coefficients for neutral
(S.sub.n), green (S.sub.G), or red (S.sub.R).
14. The method in claim 5 wherein said operational settings determining
step comprises the step of determining, in response to differences in the
measured data values for corresponding portions of said proof image and
the target image, changes in process color solid area density values and
halftone dot size through use of the following equations:
##EQU14##
where: R.sub.t.sup.T and G.sub.t.sup.T respectively represent "t"
coordinate values for red and green solid areas for the target image;
R.sub.t.sup.P and G.sub.t.sup.P respectively represent "t" coordinate
values for red and green solid areas for the proof image;
R.sub.d.sup.T and G.sub.d.sup.T respectively represent "d" coordinate
values for red and green solid areas for the target image;
R.sub.d.sup.P and G.sub.d.sup.P respectively represent "d" coordinate
values for red and green solid areas for the proof image;
DotArea represents a resulting change in dot area;
S.sub.area represents a sensitivity coefficient; and
D.sub.tint represents a desired change in tint density.
15. Apparatus for use in conjunction with first and second different
imaging systems, wherein said first and second imaging systems generate
respective first and second image depictions of a common original image,
and said second imaging system has a variable response, apparatus for
obtaining an appearance match between said first and second image
depictions and for generating operational settings for varying the
response of said second imaging system so as to calibrate a response of
the second imaging system to a response of the first imaging system, the
apparatus comprising:
means for obtaining data values for corresponding portions of said first
and second image depictions in a pre-determined manner that substantially
parallels interpretative color preferences of a viewer; and
means for determining, in response to said data values and through a
pre-defined model of said second imaging system and pre-determined
matching principles, operational settings for setting the response of said
second imaging system to produce an image depiction, through said second
imaging system, that is an appearance match to said first image depiction,
wherein said matching principles comprise a plurality of pre-defined rules
that collectively define an appearance match between corresponding image
depictions produced by said first and second imaging systems in response
to a common source image applied as a common input to both said first and
second imaging systems.
16. The apparatus in claim 15 wherein said data values obtaining means
comprises:
means for obtaining measurement data for the corresponding portions of said
first and second image depictions; and
means for transforming said measurement data into a pre-defined coordinate
system that represents color information in the predetermined manner to
yield said data values.
17. The apparatus in claim 16 wherein said first imaging system comprises
either a press or a first proofing system and said first image depiction
comprises a target image; and said second imaging system comprises a
second proofing system and said second image depiction comprises a proof
image.
18. The apparatus in claim 17 wherein said measurement data transforming
means comprises means for converting each measurement triplet r-g-b (red,
green, blue) value from r-g-b coordinate space into a corresponding a-t-d
triplet value.
19. The apparatus in claim 18 wherein said converting means utilizes the
following equation:
##EQU15##
20. The apparatus in claim 17 wherein said pre-determined matching
principles comprise:
(a) for cyan, yellow, and magenta solid process colors:
first matching an "a" coordinate associated with ones of said data values
for corresponding three-color solid overprint test patches in both the
proof image and the target image;
second matching, in a "t-d" plane, a hue angle associated with ones of said
data values for corresponding solid red test patches in both the proof
image and the target image;
third matching, in a "t-d" plane, a hue angle associated with ones of said
data values for corresponding solid green test patches in both the proof
image and the target image; and
(b) for three-color tint overprints:
fourth matching tint densities in either a-t-d or r-g-b coordinates at a
given dot size for data values associated with corresponding three-color
tint overprint test patches in both the proof image and the target image;
and
(c) for solid black and black tint;
fifth matching an "a" coordinate value in terms of "N-channel" density for
data values associated with corresponding solid black and black tint test
patches in both the proof image and the target image.
21. The apparatus in claim 17 wherein the pre-defined model is a linearized
model of the response of said second proofing system at a given operating
point, said model comprising a plurality of sensitivity coefficients which
collectively define changes in expected response of said second proofing
system to pre-defined changes in said operating point.
22. The apparatus in claim 21 wherein said changes in said operating point
comprise changes in solid area density values of cyan, yellow, magenta or
black process colors or halftone dot sizes associated with any one of said
process colors.
23. The apparatus in claim 22 wherein said pre-defined model is given by
the following equations:
(a) for three-color (red, green and blue) solid overprints:
##EQU16##
where: C.sub.r represents the change in cyan process color solid area
density for red light;
M.sub.g represents the change in magenta process color solid area density
for green ;
Y.sub.b represents the change in yellow process color solid area density
for blue; and
N.sub.a represents a resulting change in an "a" coordinate value for a
three color overprint;
(b) for red and green solid colors:
##EQU17##
where: G and R represent resulting changes in green and red coordinates
(with subscripts "t" and "d" denoting the "t" and "d" coordinates,
respectively);
(c) for three-color tint overprints
##EQU18##
and (d) for black solids and tints:
.DELTA.K.sub.a =S.sub.n .DELTA.K.sub.n
where:
K represents a resulting change in an "a" coordinate value for black (K)
solid or black tint density; and
S represents a corresponding matrix of sensitivity coefficients for neutral
(S.sub.n), green (S.sub.G), or red (S.sub.R).
24. The apparatus in claim 17 wherein said operational settings determining
means comprises means for determining, in response to differences in the
measured data values for corresponding portions of said proof image and
the target image, changes in process color solid area density values and
halftone dot size through use of the following equations:
##EQU19##
where: R.sub.t.sup.T and G.sub.t.sup.T respectively represent "t"
coordinate values for red and green solid areas for the target image;
R.sub.t.sup.P and G.sub.t.sup.P respectively represent "t" coordinate
values for red and green solid areas for the proof image;
R.sub.d.sup.T and G.sub.d.sup.T respectively represent "d" coordinate
values for red and green solid areas for the target image;
R.sub.d.sup.P and G.sub.d.sup.P respectively represent "d" coordinate
values for red and green solid areas for the proof image;
DotArea represents a resulting change in dot area;
S.sub.area represents a sensitivity coefficient; and
D.sub.tint represents a desired change in tint density. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
TECHNICAL FIELD OF THE INVENTION
The invention relates to a technique, specifically apparatus and associated
methods employed therein, for objectively providing an accurate appearance
match between a image produced by one imaging system (i.e. a "target"
image produced by, e.g. a printing press) to the same image produced by a
different imaging system (i.e. a "replica" of the target image but
produced by, e.g., a halftone color proofing system) and thereby calibrate
the performance of the latter system to that of the former system.
BACKGROUND ART
Currently, color images are generated through a wide variety of different
systems, such as for example photographically on suitable film or
photosensitive paper, or electronically on video tape or other suitable
media. When generated, images share a basic characteristic: they are
recorded on a continuous tone (hereinafter referred to as "contone")
basis. As such, recorded color information at any point in the image is
represented by several continuous amplitude values, each of which is
oftentimes discretized as eight-bit values ranging from "0" to "255". Very
often, a user having an image captured on one medium, such as a
photographic print or transparency, will desire to display and/or
reproduce that image on other media, such as on a video monitor or on a
printed page.
Color reproduction equipment, as it relates to printing images, takes
advantage of the principle that the vast majority of colors can be
separated into a specific combination of four primary subtractive colors
(cyan, yellow, magenta and black--C, Y, M and K) in which the amount of
each primary color is set to a predetermined amount. In the case of
printed reproductions of an image, use of primary color printing obviates
the need to use a differently colored ink for each different color in the
image. As such, each image is commonly converted into sets of three or
four color separations, in which each separation is essentially a negative
(or positive) transparency with an altered tone reproducing characteristic
that carries the color information for only one of the primary colors.
Separations are subsequently recorded on printing plates for use in a
press.
By way of contrast, color reproduction on cathode ray tube displays takes
advantage of the principle that the vast majority of colors can be
represented by a combination of three primary additive colors
(specifically red, green and blue--R, G and B) in which the intensity
produced by each primary colored (R, G or B) phosphor is set to a
predetermined amount.
Modern offset printing presses do not possess the capability of applying
differential amounts of ink to any location in an image being printed.
Rather, these presses are only designed to either apply or not apply a
single amount of ink to any given location on a page. Therefore, an offset
printing press is unable to directly print a contone separation. To
successfully circumvent this problem, halftone separations are used
instead. An image formed from any single color halftone separation encodes
the density information inherent in a color image from amplitude modulated
form into a spatial (area) modulated form, in terms of dot size, which is
subsequently integrated by the human eye into a desired color. By smoothly
changing halftone dot sizes (dot areas), smooth corresponding tone
variations will be generated in the reproduced image. Given this, the art
has taught for some time that a full color image can be formed by properly
overlaying single color halftone reproductions for all of the primary
subtractive colors, where each reproduction is formed from a corresponding
halftone dot separation that contains dots of appropriate sizes. Clearly,
as size and spacing of the dots decrease, an increasing amount of detail
can be encoded in a halftone dot pattern and hence in the reproduced
image. For that reason, in graphic arts applications, a halftone
separation utilizes closely spaced dots to yield a relatively high
resolution.
With this in mind, one might first think that printing a color image for
graphic arts use should be a fairly simple process. Specifically, a color
image could first be converted into corresponding continuous tone
separations. Each of these contone separations could then be converted
into a corresponding halftone separation. A printing plate could then be
manufactured from each halftone separation and subsequently mounted to a
printing press. Thereafter, paper or other similar media could be run
through the press in such a fashion so as to produce properly registered
superimposed halftone images for all the subtractive primary colors
thereby generating a full color reproduction of the original image.
In practice, accurately printing a color image is oftentimes a very
tedious, problematic and time consuming manual process that requires a
substantial level of skill. First, the conventional manual photographic
process of converting a contone separation into a halftone separation,
this process commonly being referred to as "screening", is a time and
resource consuming process in and of itself. Second, various phenomena,
each of which disadvantageously degrades an image, often occur in a
reproduced halftone color image. Moreover, the complete extent to which
each of these phenomena is present in the reproduced image is often known
only at a rather late point in the printing process thereby necessitating
the use of tedious and time and resource consuming iterative
experimentation to adequately eliminate these phenomena.
Traditionally, on-press proofing provided the first point at which a color
judgment could be made regarding the quality of the reproduced image. For
example, many color differences, such as incompatible and/or objectionable
color renditions or Moire patterns, were usually first seen at this point
in an imaging process. If such a difference were sufficiently
objectionable to a color technician, then usually the entire imaging
process would need to be modified and repeated. Doing so generally
necessitated a total rework of the separations, production of a new set of
printing plates therefrom and generation of a new press proof, with this
process being iteratively repeated as many times as necessary to properly
remove or sufficiently attenuate the incompatible and/or objectionable
color differences.
In an effort to reduce the time required and expense associated with
conventional manual photographic based color reproduction processes and
particularly the traditional on-press proofing techniques used therewith,
the art has turned away from use of on-press proofing in high volume
graphic art applications and towards the use of intermediate off-press
proofing technologies, such as electro-photographic techniques. In this
regard, U.S. Pat. No. 4,708,459 (issued to C. Cowan et al on Nov. 24,
1987, assigned to the present assignee hereof and hereinafter referred to
as the U.S. Pat. No. '459 Cowan et al patent) discloses an
electro-photographic color proofing system (also referred to herein as a
"proofer") with variable tone reproduction characteristics.
While the proofing system described in the U.S. Pat. No. '459 Cowan et al
patent provides an excellent quality proof, this system, like all imaging
systems, can reproduce colors only within a certain color gamut. Generally
speaking, the tone reproduction characteristics of one type of imaging
system, or even one type of imaging medium, are not completely coincident
with those of a different type of imaging system or medium. In this regard
through use of differing colorants (inks used in printing as compared to
photographic dyes or colored phosphors on a video monitor) and other
physical phenomena related to specific imaging processes, a given color
shown on a color artwork, such as on a photograph, or on a press sheet
printed on publication stock will often appear differently in a halftoned
color proof formed on electro-photographic film and subsequently
transferred to paper, the latter having characteristics similar to paper
which will be used in a press. Furthermore, a halftone color proofing
system, such as that described in the U.S. Pat. No. '459 Cowan et al
patent, is generally incapable of producing the exact same color gamut and
color response which are available through either the photograph or press
sheet. In this regard, the color gamut reproducible in a color halftoned
proof will generally not match that associated with a color artwork that
appears on a photograph or on a press sheet. In addition and owing to
physical differences among different imaging systems, the response of
different types of imaging systems to an identical input color will likely
be different, e.g. the same red color provided as input to two different
imaging systems might likely produce two output colors with somewhat
differing red hues.
In view of the inherent tone and color differences between, e.g., the press
sheet and the proof thereof, the colors in the proof can not be
identically matched to those that appear in the press sheet. Nevertheless,
for a proofing system to fully serve its intended purpose, a proof image
must accurately predict the image as subsequently printed on a press
sheet. However, the tone and color reproduction characteristics of a
proofing system rarely coincide with those of an associated press.
Therefore, the tone and color reproduction characteristics of the proofing
system must be calibrated, to the extent possible, to those of the press.
Once calibrated, the proofing system should be able to accurately predict
the performance of the press though, in most situations, it will generate
a proof image with colors that do not exactly match those in the press
sheet.
Unfortunately, calibrating a proofing system tends to consume an inordinate
amount of time as well as require a very high level of skill. In this
regard, a color technician is required to possess a substantial level of
skill and expertise not only to judge color differences between a proof
and a press sheet therefor but also to fully appreciate the performance
inter-relationships between the colors that appear on the proof and the
corresponding ones that will appear on the press sheet. Consequently, the
technician not only must recognize a color difference and decide which
specific colors to match but also, where the tone and color reproduction
characteristics of the proofing system can be varied, determine the proper
variations in these characteristics in order to achieve an acceptable
match between the proof and the press sheet and then set the proofing
system accordingly.
In particular, to calibrate a proofing system to a press, a color
technician usually visually examines both a proof and the associated press
sheet on a side-by-side basis and then, based upon his own subjective
judgment as to what the visually important features of the press sheet are
and how they should appear, selects which colors to match. Thereafter,
given his knowledge of the response of the proofing system and its color
response, he will attempt to initially vary the C, Y, M and K colorant
solid area densities and/or dot size (tone reproduction curve) settings to
accurately depict one color(s), which, not surprisingly, will also affect
other colors, possibly adversely. Based upon the effects that occur with
respect to other colors in the proof, the technician will iteratively vary
the solid area densities and/or dot size (tone reproduction curve)
settings of the colorants, in seriatim, until an acceptable color match is
achieved between the press sheet and the proof for the selected colors.
However, a proofing system with variable tone and color reproduction
characteristics often presents the technician with an enormous number of
different possible combinations of the settings. For example, for the
system described in the U.S. Pat. No. '459 Cowan et al patent, the solid
area density and dot size can be set for each of the four process colors
(C, Y, M and K) at any of 20 different density levels and at any of 15
different dot size settings. In view of the resulting huge number of
potential combinations of settings, an experienced color technician often
needs to run and separately analyze quite a few successive proofs in order
to select a suitable solid area density and halftone dot size setting for
each different colorant in order to achieve an acceptable match between
the proof and a press sheet and thereby calibrate the proofing system to
the press. Moreover, additional time is consumed whenever the technician
is forced to resort to trial-and-error experimentation or, in a worst case
scenario, guesswork: either merely as a result of iterating through a very
large number of possible combinations to discern the performance
inter-relationships of the proofing system and/or by incorrectly relying
on intuition and initially iterating away from a proper operating
condition. An example of the latter situation can occur where the
technician, based upon his own intuition, views a proof against a press
sheet and decides that the yellow content in the proof needs to be
increased. While the technician may decide to initially increase the
halftone dot size for the yellow colorant, the proper operating condition
may instead involve reducing the halftone dot sizes for all the colorants
but reducing the halftone dot size for yellow less than that for each of
the other colorants.
Furthermore with certain images, the technician may simply have
insufficient skill to quickly determine the proper operating conditions of
the proofing system. As such, in certain situations, the technician, given
his lack of knowledge or experience, may be unable to determine the best
possible color match in the time allotted and thus must settle for one
that is often simply acceptable. In view of this, empirical approaches
have been developed to aid the technician in quickly locating a limited
region of the operating space of the proofing system in which a decent
match can be achieved to which the proofing system can be calibrated. One
such empirical approach could involve first matching the C, M, Y and K
solid area and halftone densities between the press sheet and the proof to
the extent realistically possible--though this may generally produce
mis-matches in overprint colors, e.g. the reds, greens and blues. Once
these primary color matches are achieved, the resulting proof is then
visually examined to determine how certain overprint colors appear, e.g.
whether the gray tones are the same as on the artwork or are too red. If
the latter occurs, then the colorants are appropriately changed, possibly
through successive iterative changes, to increase the cyan content or
decrease the magenta and yellow content in the proof. Alternatively, the
technician could visually examine the reds in the proof. If the reds
appear too orange, the colorants could be appropriately changed to
decrease the yellow content of the proof or alternatively increase its
magenta content. In that regard, it is widely known that an average human
vision is acutely sensitive to flesh tones (which specifically contain red
hues). Hence, even a subtle difference in coloration may be perceived as
transforming an otherwise pleasant image of a human face into one that is
quite unnatural and obnoxious. Through such approaches, even a skilled
color technician may still need to generate upwards of 12-15 separate
proofs in seriatim, typically requiring a full day of work, until he
discerns the proper operating condition of the proofing system which is
needed to achieve an acceptable color match between the proof and a press
sheet therefor and thereby calibrate the proofing system to the press in
use.
Through a totally different approach, the technician could quantitatively
measure reflection densities of selected portions of the image on both the
press sheet and the proof using, for example, a reflection densitometer,
and then attempt to set the colorants in a manner that seeks to achieve
the densities inherent in the press sheet. Unfortunately, this approach is
constrained by the ability of the technician to locate corresponding
relatively large uniformly colored areas on both the press sheet and the
proof at which the reflection densitometer can be reliably placed to take
measurements. If both images contain significant detail, then suitable
measurement areas may not exist and thereby preclude such densitometric
measurements from being made.
Apart from a reflection densitometer, one device that has recently become
available for color measurement and matching is a spectrophotometer, such
as the Model SPM 50 spectrophotometer manufactured by Gretag Corporation
of Bothell, Wash. This device projects white light onto an image and then
separates the spectrum of reflected light from the image through a
diffraction grating and measures the intensity of the reflected radiation
at a number of different wavelengths. Through this device and its
associated software, colorimetric spectral based measurements can be made
of any reflection image. Though this device is intended to be used to
determine proper halftone dot size in the separations in order to achieve
a desired coloration in the reflection image therefrom, conceivably it
could be used to characterize (i.e. "model") a proofing system in use and
then effectuate a color balance between a press sheet and a proof
therefor. Specifically, a set of known test (reference or calibration)
separations is provided with the device, and dot area settings for these
separations are stored within the device. To characterize the proofing
system, a proof is made from the reference separations. Thereafter,
spectrophotometric measurements are taken of this particular proof. The
resulting measurements, when processed, would yield a model that
characterizes the color gamut producible through the proofing system.
Thereafter, in order to generate a color match to a press sheet, the
device could then be used to take spectrophotometric measurements of the
press sheet. Given the characterization of the proofing system and the
latter set of measurements, the software will determine appropriate values
to use for solid area densities and corresponding halftone dot sizes for
each primary colorant in the proofing system in order to generate a proof
that should match the press sheet.
Inasmuch as the color gamut reproducible through a proofer does not
coincide with that appearing in the press sheet, the software used with
this device is constrained, just as the technician is in manually
performing a color match, to effectuate a compromise in matching the two
gamuts between the press sheet and the proof therefor. In achieving a
color match, this software relies on the well-known CIELAB L*,a*,b* color
coordinate system and color differences associated therewith. In computing
a color match, the software seeks to minimize an overall .DELTA.E value
(i.e. the sum of the squares of the CIELAB color difference values)
between the two color gamuts and thus obtain a "colorimetric" match.
In this regard, a very small colorimetric difference for some colors will
lead to a very large .DELTA.E value; while this will not be true for other
colors. Any system, such as the Gretag spectrophotometer, that seeks to
minimize an overall colorimetric error between two images produced by
systems with differing tone and color reproduction characteristics may
well still produce minor color mismatches for some colors that, in various
image contexts, would be highly objectionable to a human observer.
In particular, it has been known for some time that human color perception,
including mental judgment, exhibits differing sensitivities for different
colors. Given this, human observers will be much more acutely aware of
what would amount to minor color differences, such as differences in
so-called "memory" colors (e.g. greens and flesh tones), in certain
pictorial contexts than in others. Accordingly, a color difference that
would simply be noticeable, if at all, in some contexts would be highly
objectionable in others. For example, people are acutely aware of very
small differences in flesh tones. A viewer will likely object to a human
face that appears too blue or green, while merely noticing, if at all, and
certainly not objecting to a tablecloth or blanket that exh | | |
