METHOD FOR TRANSFORMING COLOR SPACES

Description

[0001] The invention relates to methods for transforming color values ​​of a first device-dependent color space into the color values ​​of a second device-dependent color space. The method is used, for example, in the printing industry to align visual color impressions.

[0002] In the printing industry, printing templates are created for printed pages, containing all elements to be printed, such as text, graphics, and images. When these templates are produced electronically, they exist as digital data. For an image, this data is generated, for example, by scanning the image point by point and line by line, breaking each pixel down into its color components, and then digitizing these components. Typically, images are scanned into the color components red, green, and blue (R, G, B), i.e., the components of a three-dimensional color space (RGB color space). However, for color printing, such as in printing presses, different color components are required. In four-color printing, these are the process colors cyan, magenta, yellow, and black (C, M, Y, K), i.e., the components of a four-dimensional color space (CMYK color space).

[0003] During the printing process, the image data must be transformed from the scanner's RGB color space to the CMYK color space of the printing process being used. If the CMYK image data generated for a printing process is to be displayed on a screen (e.g., LCD) as a so-called soft proof, or if the image data is to be output as a proof beforehand, for example, on an inkjet printer, further color transformations are required to ensure that the visual impression of the colors corresponds as closely as possible to the later printing process used for the final print run.

[0004] Such color transformations are necessary in the printing industry because all devices and processes have certain limitations and characteristics in the representation and reproduction of colors. Therefore, different color spaces exist for various devices and processes such as scanners, monitors, proofing devices, printing processes, and the like. These color spaces describe the color characteristics of the respective device or process as accurately as possible and are referred to as device-dependent color spaces.

[0005] In addition to device-dependent color spaces, there are also device-independent color spaces, which are based on the human visual characteristics of a so-called "normal" observer. Examples of such color spaces are the XYZ color space defined by the standardization organization CIE (Commission Internationale d'Éclairage) and the L*a*b* color space derived from it. The L*a*b* color components can be converted into XYZ color components and vice versa. To determine whether two colors are perceived as the same or different by the human eye under identical environmental conditions, especially the same lighting, measuring the XYZ or L*a*b* color components is sufficient. The L*a*b* color components form a color space with a luminance axis [L*] and two chromatic axes [a*, b*], which can be visualized in the plane of a color circle, with the luminance axis passing through its center.

[0006] US 7,161,710 B1 describes the transformation of color spaces by chaining rendering intents, for example for printing applications.

[0007] DE 103 22 378 A1 describes a method for color transformation between color spaces using color profiles, whereby rendering intents are used.

[0008] US 2007 / 058181 A1 describes a method for converting color values ​​between different devices, in which both relative and absolute color adjustments can be made using conversion rules.

[0009] A device or color processing process can be characterized with respect to its color properties by assigning the XYZ color components that a human sees in the colors produced with these combinations to all possible value combinations of the associated device-dependent color space. For a printing process, the different CMYK value combinations each produce a different printed color. A colorimeter can be used to determine the XYZ components of the printed colors and assign them to the CMYK value combinations. Such an assignment, which relates the device-dependent colors produced by a device or process to a device-independent color space (XYZ or L*a*b*), is called a color profile; in the case of a printing process, it is called an output color profile. The definitions and data formats for color profiles have been standardized by the ICC (International Color Consortium).An ICC color profile stores the mapping of color spaces in both directions, for example, the mapping XYZ = f1 (CMYK) and the inverse mapping CMYK = f2 (XYZ). The mapping defined by a color profile can be implemented using a lookup table. For example, if the CMYK color components of a printing process are to be mapped to the XYZ color components, the lookup table must have a memory location for each possible combination of CMYK color values, in which the mapped XYZ color components are stored. However, this simple mapping method has the disadvantage that the lookup table can become very large, so interpolation methods are generally used.

[0010] The mappings between device-dependent color spaces and a device-independent color space given in the color profiles can be used for color transformation between the device-dependent color spaces, so that, for example, the color values ​​[C1, M1, Y1, K1] of a first printing process are converted accordingly into the color values ​​[C2, M2, Y2, K2] of a second printing process, so that the second print has the same colors as the first print according to the visual impression.

[0011] Figur 1 Figure 1 schematically illustrates the principle of such a color transformation for a state-of-the-art printing process adaptation. A first color transformation 1 from the color values ​​[C1, M1, Y1, K1] of the first printing process into XYZ color values ​​and a second color transformation 2 from the XYZ color values ​​into the color values ​​[C2, M2, Y2, K2] of the second printing process are performed sequentially. The two color transformations 1 and 2 can also be combined into an equivalent color transformation 3, which directly maps the color values ​​[C1, M1, Y1, K1] and the color values ​​[C2, M2, Y2, K2] to each other. Since the color values ​​[C1, M1, Y1, K1] and [C2, M2, Y2, K2] are assigned to each other via the device-independent XYZ intermediate color space, which result in the same XYZ color values, the assigned printing inks in the two printing processes are largely perceived as visually identical within the printing color gamut.

[0012] In the ICC specification, the device-independent color space used to link device-dependent color spaces during color transformation is called the Profile Connection Space (PCS). The Profile Connection Space is the interface between the color profiles of devices and processes. It is defined as an ideal reference image color space in an ideal viewing environment. It is based on the standard color spaces CIE 1931 XYZ and CIE 1976 L*a*b* defined by the CIE. The white point of the Profile Connection Space is defined by the standard illuminant D50 commonly used in graphic design, i.e., illumination with a light source of 5000 Kelvin. This white point, WPD50, has the following XYZ color values: X WPD 50 = 0 , 9642 Y WPD 50 = 1 , 0000 Z WPD 50 = 0 , 8249

[0013] The mappings between a device-dependent color space and the Profile Connection Space described in the ICC profiles exist in various forms, depending on the rendering intent. These rendering intents are referred to as "Relative Colorimetric" or "Relative," "Absolute Colorimetric" or "Absolute," "Perceptual," and "Saturation." They differ, among other things, in the gamut mapping built into the color profiles. Gamut mapping refers to the method or strategy used to adapt the different color gamuts of the device-dependent color spaces to one another. For example, not all bright and saturated colors that can be displayed on a monitor are printable, especially when printing on lower-quality, relatively gray paper, such as newsprint.Then, by assigning the color profile, the non-printable monitor colors must be converted into similar colors on the edge of the color gamut of the printable colors, so that an overall harmonious color impression is created without subjectively perceived color distortions.

[0014] The Perceptual Rendering Intent aims to consider not only visually perceived color accuracy but also other properties important for image reproduction, such as contrast, detail, and the real viewing environment, when mapping the image to the Profile Connection Space. The Saturation Rendering Intent primarily preserves pure and saturated colors and is used for reproducing graphics.

[0015] The relative rendering intent is used, for example, to map different printing processes to each other using a color transformation, fully utilizing the color gamut and brightness range of the target process. Specifically, this means that the white point of the source process, i.e., the white of the paper, is mapped to the white point of the target process. If the white point of the target process is brighter than the white point of the source process, the colors will be rendered brighter and more brilliantly after the color transformation when printed with the target process.

[0016] With absolute rendering intent, on the other hand, the white point and XYZ color values ​​of the source process are reproduced unchanged during color transformation when printing with the target process. This requires that the printable color gamut and brightness range of the target process are larger than those of the source process. Absolute rendering intent is therefore used to reproduce a source print process as a color-accurate and binding proof with a target print process, for example, on a high-quality inkjet printer.

[0017] In the mapping tables created for the relative rendering intent, the XYZ color values ​​of the Profile Connection Space assigned to the device-dependent color values ​​are scaled to fully utilize the possible value range of the Profile Connection Space. Specifically, this means that the measured white point WP1 of the device-dependent color space (Media White Point) is assigned the white point WPD50 in the Profile Connection Space. If the white point WP1 of a source printing process, for example newspaper printing, has the measured XYZ color values ​​[X WP1 , Y WP1 , Z WP1 ], then when creating the color profile, all color values ​​[X1, Y1, Z1] of the color patches measured for different value combinations [C1, M1, Y1, K1] on a test template are scaled component-wise by the ratio of the white points WPD50 and WP1 to obtain the assigned color values ​​[X PCS1 , Y PCS1 , Z PCS1 ] of the Profile Connection Space. X PCS 1 = X 1 × X WPD 50 / X WP 1 Y PCS 1 = Y 1 × Y WPD 50 / Y WP 1 Z PCS 1 = Z 1 × Z WPD 50 / Z WP 1

[0018] Similarly, when creating the color profile for a target printing process with the white point WP2, for example offset printing, the color values ​​[X2, Y2, Z2] measured for different value combinations [C2, M2, Y2, K2] are scaled component-wise by the ratio of the white points WPD50 and WP2 to obtain the assigned color values ​​[X PCS2 , Y PCS2 , Z PCS2 ] of the Profile Connection Space. X PCS 2 = X 2 × X WPD 50 / X WP 2 Y PCS 2 = Y 2 × Y WPD 50 / Y WP 2 Z PCS 2 = Z 2 × Z WPD 50 / Z WP 2

[0019] Since when linking the color profiles according Figur 1 When identical value combinations [X PCS1 , Y PCS1 , Z PCS1 ] and [X PCS2 , Y PCS2 , Z PCS2 ] are assigned to each other in the Profile Connection Space, the following relationship results during the color transformation of the source process to the target process according to the relative rendering intent: X 2 = X 1 × X WP 2 / X WP 1 Y 2 = Y 1 × Y WP 2 / Y WP 1 Z 2 = Z 1 × Z WP 2 / Z WP 1

[0020] The device-dependent color values ​​[C1, M1, Y1, K1] of the source process are thus transformed into the device-dependent color values ​​[C2, M2, Y2, K2] of the target process such that the corresponding XYZ color values ​​are scaled component-wise in proportion to the white point values. In particular, it follows from equation (3) that the white point WP1 of the source process is transformed into the white point WP2 of the target process.

[0021] Such a simple scaling of the XYZ color values, as determined by the ICC specification for relative rendering intent, is not optimal when the white points of the source and target processes are relatively far apart. In this case, despite the linear scaling of the XYZ color values ​​in the target process, the relative distances of the colors printed on the media with the different white points are not perceived as equivalent to the source process, because the human visual system performs a white point-dependent chromatic adaptation when viewing the colors of the target printing process.

[0022] However, if no transformation of color values ​​takes place between different devices, then when outputting to a different output device, color values ​​will be altered for a different paper white in the output process compared to the paper white of the input process. Accordingly, a light red that was originally printed on a light substrate will appear darker when the same image is output (e.g., as a PDF file) on a darker substrate.

[0023] The methods known from the prior art for converting color values ​​have disadvantages. In particular, it is desirable to perform the conversion of color values ​​without having to simulate the color of the substrate (paper white).

[0024] The object of the present invention is to provide a method by which an effective conversion of color values ​​between two different device-dependent color spaces is possible and with which one obtains the same visual impression in both color spaces.

[0025] Accordingly, the present invention relates to a method for transforming color values ​​according to claim 1.

[0026] According to the invention, at least one absolute rendering intent is combined with at least one relative rendering intent. In a preferred embodiment, "combined" means that the mappings of the two rendering intents, absolute rendering intent and relative rendering intent, are blended. In other words, in some areas of the color space, the absolute colorimetric mapping (the absolute rendering intent) is used, while in other areas of the color space, particularly for the color of the substrate background, preferably paper white, the relative colorimetric mapping (the relative rendering intent) is used, and a blend is performed between them. Since the transformations [C1,M1,Y1,K1]=>[X,Y,Z] and [X,Y,Z]=>[C2,M2,Y2,K2] (see Figur 1 ) are not linear and furthermore, if blending in [X,Y,Z], or more precisely the color space coordinates after linearized Bradford transformation, works with an exponent (see below), then [C2,M2,Y2,K2] = a*[C2,M2,Y2,K2] absolute + (1-a)*[C2,M2,Y2,K2] relative only approximates this, where the blending parameter a depends on the position in the color space.

[0027] In a particularly preferred embodiment, the combination of absolute rendering intent and relative rendering intent according to the invention means that when transforming the color values ​​of the first device-dependent color space into the color values ​​of the second device-dependent color space, a portion of the color values ​​of the first device-dependent color space are transformed by means of at least one absolute rendering intent and a portion of the color values ​​of the first device-dependent color space are transformed by means of at least one relative rendering intent.

[0028] In a preferred embodiment, the at least one absolute rendering intent and the at least one relative rendering intent are each colorimetric rendering intents.

[0029] In another preferred embodiment, when transforming the color values ​​of the first device-dependent color space into the color values ​​of the second device-dependent color space, exactly one absolute rendering intent is combined with exactly one relative rendering intent.

[0030] According to the invention, the transformation of the color values ​​of the first device-dependent color space into the color values ​​of the second device-dependent color space is carried out via the color values ​​of an intermediate color space. In a preferred embodiment, the intermediate color space is a device-independent color space (Profile Connection Space). In a particularly preferred embodiment, the color values ​​of the intermediate color space correspond to the D50 2° standard observer.

[0031] According to the invention, the color values ​​of the first device-dependent color space are transformed into the color values ​​of the intermediate color space, and the color values ​​of the intermediate color space are transformed into the color values ​​of the second device-dependent color space. According to the invention, when transforming the color values ​​of the first device-dependent color space into the color values ​​of the intermediate color space, at least one absolute rendering intent is combined with at least one relative rendering intent. According to the invention, when transforming the color values ​​of the intermediate color space into the color values ​​of the second device-dependent color space, at least one absolute rendering intent is combined with at least one relative rendering intent.According to the invention, both in the transformation of the color values ​​into the intermediate color space and in the transformation of the color values ​​from the intermediate color space into the color values ​​of the second device-dependent color space, at least one absolute rendering intent is combined with at least one relative rendering intent.

[0032] Standard ICC profiles contain three conversion tables for four rendering intents: i) one for perceptual, ii) one for saturation, and iii) one for both absolute and relative rendering. In a preferred embodiment, rendering intents are not combined from two different conversion tables, but only within the same conversion table. Accordingly, for example, the combination of perceptual and absolute rendering intents is not preferred.

[0033] Accordingly, in a preferred embodiment, the color values ​​of the first device-dependent color space are transformed into the color values ​​of the intermediate color space via at least one profile comprising at least one conversion table, and the color values ​​of the intermediate color space are transformed into the color values ​​of the second device-dependent color space via at least one profile comprising at least one conversion table, wherein the at least one absolute rendering intent and the at least one relative rendering intent use the same conversion table. In a particularly preferred embodiment, the at least one profile is at least one ICC profile. The intermediate color space is then preferably the Profile Connection Space (PCS).

[0034] In a particularly preferred embodiment, the color values ​​of the first device-dependent color space are transformed into the color values ​​of the intermediate color space via at least one profile comprising at least one conversion table, wherein the absolute rendering intent and the relative rendering intent use the same conversion table. In a further particularly preferred embodiment, the color values ​​of the intermediate color space are transformed into the color values ​​of the second device-dependent color space via at least one profile comprising at least one conversion table, wherein the absolute rendering intent and the relative rendering intent use the same conversion table.It is also possible to use the inventive combination of absolute rendering intent and relative rendering intent via at least one profile with at least one conversion table simultaneously both in the transformation into the color values ​​of the intermediate color space and in the transformation out of the color values ​​of the intermediate color space.

[0035] In another preferred embodiment, the device to which the first device-dependent color space refers is a different device than the device to which the second device-dependent color space refers.

[0036] In a further preferred embodiment, the transformation of the color values ​​of the first device-dependent color space into the color values ​​of the second device-dependent color space is carried out during a printing process in which a digital print image is printed onto a substrate using a printing press, wherein the first device-dependent color space is the color space of a screen and the second device-dependent color space is the color space of the printing press. In a particularly preferred embodiment, the printing process is selected from letterpress printing, planographic printing, gravure printing, and combinations thereof, in particular offset printing, inkjet printing, flexographic printing, screen printing, and gravure printing.

[0037] In another preferred embodiment, no simulation of color values ​​of the unprinted substrate is performed when transforming the color values ​​of the first device-dependent color space into the color values ​​of the second device-dependent color space.

[0038] In another preferred embodiment, the color value of the unprinted substrate in the first device-dependent color space is mapped to the color value of the unprinted substrate in the second device-dependent color space.

[0039] In another preferred embodiment, the color values ​​of the unprinted substrate are paper white.

[0040] In a further preferred embodiment, the transformation of the color values ​​of the first device-dependent color space into the color values ​​of the second device-dependent color space transforms the brightest point of the first device-dependent color space into the brightest point of the second device-dependent color space, resulting in an output without dots. In other words, the transformation is such that a color value corresponding to the white point in the input color space—for example, unprinted paper with a printer color space as the input color space—is also mapped to the white point in the output color space. This means that for such a color value, no ink is printed onto the substrate in the output. In other words, unprinted areas in the input remain unprinted areas in the output, and no dots are printed to simulate the darker white of the paper in the input.

[0041] In a further preferred embodiment, the at least one relative rendering intent includes black point compensation. This means that the darkest point of the input process is also mapped to the darkest point of the output process. Such compensation ensures that depth detail is preserved. Without such compensation, all colors that are darker than the darkest color in the output color space after relative / absolute / combined color transformation are clipped in the output process and attached to the lower boundary of the representable color space. This can result in clipping in gradients.

[0042] The present invention also relates to a printing press with which the inventive method is carried out. Preferred printing presses on which the inventive method is carried out are offset printing presses, inkjet printing presses, flexographic printing presses, screen printing presses and gravure printing presses.

[0043] The present invention also relates to the use of the method according to the invention in a printing press. Preferred printing presses in which the method according to the invention is used are offset printing presses, inkjet printing presses, flexographic printing presses, screen printing presses and gravure printing presses.

[0044] The rendering intents used in the prior art so far aim for either a pleasing (photographic and photographic, saturation-preserving and relative colorimetric rendering intent) or an absolute reproduction (absolute rendering intent) of the device-specific input color space in the device-specific output color space. The rendering intents "relative colorimetric" and "relative colorimetric with shadow compensation" are considered here as simple implementations of a pleasing mapping.

[0045] By definition, the absolute rendering intent simulates the dark white point when simulating processes with dark white points on processes with a lighter white point. This is often undesirable in the packaging industry. In contrast, the other rendering intents are designed so that the entire color body of the input process is transformed into the output color body. While both variants of the relative rendering intent achieve this by simply multiplying the white point and the darkest point in XYZ, the perceptual and saturation-preserving rendering intents use the strategy of the respective ICC profile algorithm.

[0046] No implementations are known that combine absolute and relative rendering intent to the brightest point of the color body of the target process.

[0047] A known method involves using paper white with an empty separation in the color space of the target process. However, simply replacing the representation of the brightest point of the input color space in the output color space with empty separations results in aesthetically displeasing transitions in images. In particular, transitions near the brightest point of the input color space, for example in wedding dresses or white blouses, appear discontinuous in the target color space at the boundary between white and a separation that is no longer filled with empty separations. To date, no method is known that can be used with standard ICC profiles.

[0048] In a preferred embodiment of the invention, template data for a substrate with a white point are to be used W 1 on a substrate with a significantly different white point WTwo outputs will be generated. ICC profiles exist for both processes, describing them. The two known colorimetric rendering intents are "absolute rendering intent" and "relative rendering intent".

[0049] Absolute Rendering Intent: Exact reproduction of the measured color values ​​(XYZ, L*a*b*) wherever the target process allows. x 2 y 2 z 2 = x 1 y 1 z 1

[0050] The paper whiteness of the initial process is simulated; this means that, under certain circumstances (target paper lighter than the source paper or a different color), white areas will also be printed. Light colors will be cut off if the target paper is too dark.

[0051] Relative Rendering Intent: Reproduction of the measured color values ​​(XYZ, L*a*b*) in relation to the white point of the respective process.

[0052] Color coordinates (x, y, z) are scaled according to the ratio of the coordinates of the white points: x 2 y 2 z 2 = x 1 ⋅ x white , 2 x white , 1 y 1 ⋅ y white , 2 y white , 1 z 1 ⋅ z white , 2 z white , 1

[0053] The color coordinates (x, y, z) can be, for example, XYZ or a coordinate system adapted to the sensitivity of the human eye. K In that case, the color coordinates of the input value and paper white must first be entered into the coordinate system. K The coordinates are transformed and then transformed back again. The white of the paper remains unprinted in the final process. On dark target paper, colored tones are also darkened accordingly.

[0054] However, what is often desired is a behavior where the colors are reproduced exactly as with absolute rendering intent, especially when the area is heavily printed (for example, a solid cyan or a red made from 100% magenta and 100% yellow). Towards the paper white, the behavior should correspond to relative rendering intent, so that the paper remains unprinted and detail is preserved even in the highlights.

[0055] Without ICC-based color management, such behavior can be achieved on an offset printing press by using the plates of the source paper, the paper with the different white point. W The second plate is inserted, and when setting up the machine, an attempt is made to match the solid tones and dot gains of the standard for the original paper as closely as possible. This procedure is only possible as long as the plates can be transferred. It cannot work if the target process uses more or fewer process colors than the source process, if spot colors in the source process are simulated by process colors in the target process, if the two processes differ not only in paper white but also in process-related characteristics (location of solid tones, co-printing behavior, etc.), or if the target process uses a different printing technology.

[0056] In the inventive approach, a method for the transformation is provided. x 2 y 2 z 2 = F x 1 y 1 z 1 The color coordinates are used as a combination of (4) and (5), which exhibits the desired behavior. The color coordinates x, y and z This could again be, for example, XYZ or a coordinate system adapted to the perception of the human eye.

[0057] Two extreme cases of transformation are possible: i) relative behavior regarding paper whiteness: F x white , 1 y white , 1 z white , 1 = x white , 2 y white , 2 z white , 2 = x white , 1 ⋅ x white , 2 x white , 1 1 y white , 1 ⋅ y white , 2 y white , 1 1 z white , 1 ⋅ z white , 2 z white , 1 1 ii) absolute behavior when at least one of the color coordinates is small: lim x 1 ↦ 0 F x 1 y 1 z 1 = x 1 y 1 z 1 = x 1 ⋅ x white , 2 x white , 1 0 y 1 ⋅ y white , 2 y white , 1 0 z 1 ⋅ z white , 2 z white , 1 0 lim y 1 ↦ 0 und lim z 1 ↦ 0 : analog

[0058] By combining equations (7) and (8), the transformation can be formulated as follows: F x 1 y 1 z 1 = x 1 ⋅ x white , 2 x white , 1 α x y 1 ⋅ y white , 2 y white , 1 α y z 1 ⋅ z white , 2 z white , 1 α z

[0059] The exponents α This depends on the color location, and the following must apply: α = 1 for paper white and α ↦ 0 when a color coordinate approaches zero. The exponents can be... α x , α y , α z theoretically, they could be different. In the following, the first approach will be restricted to the case where the exponents are the same for all components. α x = α y = α z = α. Thus, the transformation equation (10) becomes F x 1 y 1 z 1 = x 1 ⋅ x white , 2 x white , 1 α y 1 ⋅ y white , 2 y white , 1 α z 1 ⋅ z white , 2 z white , 1 α The challenge now is to determine the exponent. α as a function of the color coordinates x 1 , y 1 , z 1 and x white,1 , y white,1 , z white,1 to determine. In addition to the boundary conditions resulting from (7) and (8), further conditions must be satisfied: 1. α = 1 für xyz 1 = xyz white , 1 2. lim x ↦ 0 α = 0 und lim y ↦ 0 α = 0 und lim z ↦ 0 α = 0 3. α should apply to all values ​​of xyz between 0 and xyz white There are 4 between 0 and 1. α should be for each component of xyz be strictly monotonically increasing 5. α should be constant. 6. So that the transformations W 1 ↦ W 2 and W 2 ↦ WSince they are exactly inversely related, it must be irrelevant whether α is determined for the original coordinates or for the transformed coordinates. That is, α xyz 1 xyz white , 1 = α xyz 2 xyz white , 2 .

[0060] Here the abbreviation was used xyz = ( x, y, z ) introduced for the entire coordinate vector.

[0061] In order to guarantee the aforementioned condition 6 of independence from the transformation direction, one first introduces a white-point-independent intermediate color space with the coordinates xyz m = ( x m , y m , z m) one. If for each initial white point W 1, at least formally, is always mapped into this intermediate color space and from there into the coordinates for the target white point. W If 2 is mapped, independence from the direction is definitely ensured if the exponent α is determined in this space.

[0062] For the sake of simplicity, only one component will be considered in the following. x of the color space and describes the mapping into the space after x = x 1 oder x 2 initially as x = f x m x white : = f x m without the notation with the exponent α. For the function f The following conditions must now apply: 1. f 0 = 0 2. d f d x m x m = 0 = 1 3. f 1 = x white 4. d f d x m > 0 ∀ 0 ≤ x m ≤ 1

[0063] To ensure these conditions, the following approach is used for a differential equation to determine... f : d f x m d x m = x white − f x m x white ⋅ 1 1 − x m

[0064] This equation has as a general solution f x m = c 1 ⋅ x m − 1 1 x white + x white and by substituting the boundary conditions, it follows x = f x m = x white ⋅ 1 − 1 − x m 1 x white

[0065] Equation (15) can now also be applied to the direction of absolute color coordinates on the paper white. W 1 or W 2. Switch to the intermediate color space: x m = 1 − 1 − x x white x white

[0066] Now, if we substitute equation (16) for x = x 1 in equation (15) for x = x If we enter 2, we obtain the transformation equation of x 1 with x white,1 after x 2 with x white,2 : x 2 = x white , 2 ⋅ 1 − 1 − x 1 x white , 1 x white , 1 x white , 2

[0067] So that equation (17) can be expressed in the form x 2 = x 1 ⋅ x white , 2 x white , 1 α writing is possible, which results in x 2 = x 1 ⋅ x white , 2 x white , 1 α mit α = log x white , 2 x 1 ⋅ 1 − 1 − x 1 x white , 1 x white , 1 x white , 2 log x white , 2 x white , 1

[0068] Since the relationship between x 1 and x Since equation (17) is independent of the transformation direction through the construction via the intermediate color space and equation (18) is only another representation of this relationship, the independence of the transformation direction must also hold for equation (18).

[0069] Equation (18) can now be applied to the three color space coordinates x, y and zThey can be applied separately. This results in a relatively colorimetric behavior towards white and an absolutely colorimetric behavior towards black. However, this method has two further special features: 1. The exponents α differ for x, y and z, which potentially leads to disturbing color shifts. 2. As soon as one of the coordinates x, y or z If the coordinate has a value significantly different from 0, it will no longer be transformed in a way similar to an absolute colorimetric transformation.

[0070] Therefore, a representative coordinate q with the following properties is introduced: i) q transforms as x, y and z, so q 2 = q 1 ⋅ q white , 2 q white , 1 α ii) q approaches 0 as soon as at least one of the coordinates x, y or z It approaches zero.

[0071] A coordinate that satisfies these conditions is the geometric mean of the coordinates. x, y and z q = x ⋅ y ⋅ z 3 q white = x white ⋅ y white ⋅ z white 3 etc.

[0072] This allows us to determine a common exponent α Determine for all coordinates, and the final transformation equation can be summarized as follows: x 2 = x 1 ⋅ x white , 2 x white , 1 α , y 2 = y 1 ⋅ y white , 2 y white , 1 α , z 2 = z 1 ⋅ z white , 2 z white , 1 α with α = log q white , 2 q 1 ⋅ 1 − 1 − q 1 q white , 1 q white , 1 q white , 2 log q white , 2 q white , 1 und q = x ⋅ y ⋅ z 3

[0073] It should be noted that the coordinates xyz They do not necessarily correspond to XYZ. If necessary, a linear transformation, such as a Bradford matrix (known in principle from the prior art), must first be applied to the XYZ coordinates, followed by the inverse Bradford matrix. The coordinates of the white points are, of course, treated analogously in this case.

[0074] Furthermore, it should be noted that the color coordinates xyz used are always absolute, meaning they are not scaled to the white point coordinates. Since internal calculations with ICC profiles usually use color coordinates in units of the white point, this scaling must be taken into account. However, because it is a simple multiplication and the white points are known at all times, this does not pose a limitation to its applicability.

[0075] For later consideration of limiting cases, it is useful to express equations (22) and (23) in centered color space coordinates, i.e. scaled with the coordinates of the white point. x ˜ = x x white y ˜ = y y white z ˜ = z z white q ˜ = q q white to represent. By substituting (24) into (22) we obtain x ˜ 2 = x ˜ 1 ⋅ x white , 2 x white , 1 α − 1 und y ˜ , z ˜ analog .

[0076] For the exponent α this results from substituting (24) into (23) α = log 1 q ˜ 1 ⋅ q white , 2 q white , 1 ⋅ 1 − 1 − q ˜ 1 q white , 1 q white , 2 log q white , 2 q white , 1 = 1 + log 1 q ˜ 1 ⋅ 1 − 1 − q ˜ 1 q white , 1 q white , 2 log q white , 2 q white , 1 .

[0077] If we now define a α̃ for the centered color space coordinates as α ˜ = α − 1 = log 1 q ˜ 1 ⋅ 1 − 1 − q ˜ 1 q white , 1 q white , 2 log q white , 2 q white , 1 Equation (25) thus becomes x ˜ 2 = x ˜ 1 ⋅ x white , 2 x white , 1 α ˜ und y ˜ , z ˜ analog .

[0078] If you put α̃ Substituting equation (27) into equation (28), it can be further simplified by introducing new abbreviations to x ˜ 2 = x ˜ 1 ⋅ exp log x white , 1 x white , 2 log q white , 1 q white , 2 ︸ : = β x ⋅ log 1 q ˜ 1 ⋅ 1 − 1 − q ˜ 1 q white , 1 q white , 2 ︸ : = f ˜ q = x ˜ 1 ⋅ f ˜ q β x .

[0079] This allows the transformation of the centered coordinates to be written compactly as x ˜ 2 = x ˜ 1 ⋅ f ˜ q β x y ˜ 2 = y ˜ 1 ⋅ f ˜ q β y z ˜ 2 = z ˜ 1 ⋅ f ˜ q β z mitβ x = log x white , 1 x white , 2 log q white , 1 q white , 2 β y = log y white , 1 y white , 2 log q white , 1 q white , 2 β z = log z white , 1 z white , 2 log q white , 1 q white , 2 und f ˜ q = 1 q ˜ 1 ⋅ 1 − 1 − q ˜ 1 q white , 1 q white , 2

[0080] The factors β in equation (31) can already be calculated when suspending the transformation, since they depend only on the color space coordinates of the white points.

[0081] The transformation F described by equations (30), (31), (32) has the property that the transformation of white point W 1 after white point W 2 and those from Whitepoint W 2 after white point W 1. Be exactly inversely related to each other, that is to say F W 2 ↦ W 1 F W 1 ↦ W 2 xyz = xyz .

[0082] However, the result of the transformation is generally not independent of whether one transforms from white point. W 1 after white point W 3 first according to white point W 2 transformed or whether one calculates the direct transformation, that is, in general the following F W 2 ↦ W 3 F W 1 ↦ W 2 xyz ≠ F W 1 ↦ W 3 xyz

[0083] To achieve this property, the transformation can be modified such that it does not directly transform between the source and target white points, but instead always selects the white point of the illumination and observation condition D50-illuminant and 2°-observer (D50O02) defined in the ICC specification as an intermediate point. This transformation G thus results from the transformation F described in equations (30), (31), (32) as G W 1 ↦ W 2 xyz = F W D 50 O 02 ↦ W 2 F W 1 ↦ W D 50 O 02 xyz

[0084] As a rule, all centered color space coordinates should x̃ , ỹ and z̃ The values ​​lie between 0 and 1. Equation (32) is therefore only valid on this interval. However, negative values ​​or values ​​above 1 can occur under certain circumstances. This can happen, for example, when considering self-luminous colors. Another reason for irregular centered color space coordinates can be, for example, a linear approximation for a Bradford transformation to the color space coordinates when extremely hued colors are transformed. To continue the transformation for these values, colors with at least one non-positive color space coordinate should be treated absolutely colorimetrically, and colors with a product of the color space coordinates greater than 1 should be treated relatively colorimetrically.

[0085] The transformation equation (30) can thus be modified to x ˜ 2 = x ˜ 1 ⋅ x white , 1 x white , 2 : x ˜ ≤ 0 ∨ y ˜ ≤ 0 ∨ z ˜ ≤ 0 absolut x ˜ 1 : x ˜ ⋅ y ˜ ⋅ z ˜ ≥ 1 relativ x ˜ 1 ⋅ f ˜ q β x : sonst . mixed ỹ , z̃ :analog.

[0086] The presented method is based on using equation (32) to determine a factor f̃ q for scaling the geometric mean q to determine the color space coordinates and to use the exponents β x , β x , β z to divide the color space coordinates. This method initially fails precisely when genuinely different white points are involved. W 1 and W 2 have the same geometric mean, that is x white , 1 ⋅ y white , 1 ⋅ z white , 1 = x white , 2 ⋅ y white , 2 ⋅ z white , 2 and x white , 1 ≠ x white , 2 ∨ y white , 1 ≠ y white , 2 ∨ z white , 1 ≠ z white , 2

[0087] This can occur precisely when the circumstances are x white , 2 x white , 1 , y white , 2 y white , 1 , z white , 2 z white , 1 Both values ​​greater than 1 and those less than 1 can occur. Therefore, the logarithmic signs are defined as Σ x = Σ x white , 2 x white , 1 , Σ y = Σ y white , 2 y white , 1 Σ z = Σ z white , 2 z white , 1 mit Σ r = − 1 : r < 1 0 : r = 1 1 : r > 1 as well as the dominant logarithmic sign as Σ ¯ = − 1 : Σ x + Σ y + Σ z < 1 1 : sonst .

[0088] This allows the ratio of the geometric means of the white points in equations (31) and (32) to be replaced. q white , 1 q white , 2 ↦ x white , 1 x white , 2 σ x ⋅ y white , 1 y white , 2 σ y ⋅ z white , 1 z white , 2 σ z 3 where the weights σ = -1, 1, which indicate which ratios are inverse and which are direct, still need to be determined.

[0089] For the direct transformation of white point W 1 after W 2 shall be the inverse component whose logarithmic sign differs from the dominant sign.

[0090] That means, σ x = − 1 : Σ x ≠ Σ ¯ 1 : sonst . and σ y , σ z Similarly, the ratio in equation (42) can only become 1 if and only if the two white points W 1 and W 2 are identical. In this case, the entire transformation is an identity and can be omitted.

[0091] For the transformation with the intermediate step via D50O02, it can generally be assumed that the color space coordinates of the white points lie below those of D50O02. Therefore, during the transformation to D50O02, those ratios with a positive logarithmic sign are inverted. Σ have and in the transformation of D50O02 those with a negative logarithmic sign Σ : q white , 1 q D 50 O 02 ↦ x white , 1 x D 50 O 02 − Σ 1 , x ⋅ y white , 1 y D 50 O 02 − Σ 1 , y ⋅ z white , 1 z D 50 O 02 − Σ 1 , z 3 with Σ 1 , x = Σ x white , 1 x D 50 O 02 Σ 1 , y = Σ y white , 1 y D 50 O 02 Σ 1 , z = Σ z white , 1 z D 50 O 02 q D 50 O 02 q white , 2 ↦ x D 50 O 02 x white , 2 − Σ 2 , x ⋅ y D 50 O 02 y white , 2 − Σ 2 , y ⋅ z D 50 O 02 z white , 2 − Σ 2 , z 3 with Σ 2 , x = Σ x D 50 O 02 x white , 2 Σ 2 , y = Σ y D 50 O 02 y white , 2 Σ 2 , z = Σ z D 50 O 02 z white , 2

[0092] These ratios can only become 1 if the respective white point W 1 [ W 2 ] exactly corresponds to the white point of the illumination condition. In this case, the transformation of W 1 to D50O02 [or from D50O02 to W 2 ] the identical image and can be omitted.

[0093] Figur 1 Figure 1 schematically illustrates the principle of a color transformation for state-of-the-art printing process adaptation. A first color transformation 1 from the color values ​​[C1, M1, Y1, K1] of the first printing process into XYZ color values ​​and a second color transformation 2 from the XYZ color values ​​into the color values ​​[C2, M2, Y2, K2] of the second printing process are performed sequentially. The two color transformations 1 and 2 can also be combined into an equivalent color transformation 3, which directly maps the color values ​​[C1, M1, Y1, K1] and the color values ​​[C2, M2, Y2, K2] to each other. Since the color values ​​[C1, M1, Y1, K1] and [C2, M2, Y2, K2] are assigned to each other via the device-independent XYZ intermediate color space, which result in the same XYZ color values, the assigned printing inks in the two printing processes are largely perceived as visually identical within the printing color gamut.

[0094] Figur 2 Figure 1 schematically illustrates the method according to the invention. The color values ​​of a first device-dependent color space are transformed into an intermediate color space. The transformation into the intermediate color space is performed by combining absolute and relative rendering intent. The color values ​​transformed into the intermediate color space are then transformed into the color values ​​of the second device-dependent color space. The transformation from the intermediate color space is performed by combining absolute and relative rendering intent. Bezugszeichenliste

[0095] 1first color transformation 2second color transformation 3third color transformation

Claims

1. Method for transforming colour values of a first device-dependent colour space into the colour values of a second device-dependent colour space, so that the visual impression of the colours reproduced in both colour spaces is essentially the same, and wherein at least one rendering intent, which serves to describe colour values, is used in the transformation of the colour values of the first device-dependent colour space into the colour values of the second device-dependent colour space, wherein at least one absolute rendering intent, which describes absolute colour values, is combined with at least one relative rendering intent, which describes relative colour values, during the transformation of the colour values of the first device-dependent colour space into the colour values of the second device-dependent colour space, wherein the transformation of the colour values of the first device-dependent colour space into the colour values of the second device-dependent colour space is carried out via the colour values of an intermediate colour space, wherein the colour values of the first device-dependent colour space are transformed into the colour values of the intermediate colour space and the colour values of the intermediate colour space are transformed into the colour values of the second device-dependent colour space, characterised in that at least one absolute rendering intent is combined with at least one relative rendering intent during the transformation of the colour values of the first device-dependent colour space into the colour values of the intermediate colour space, and in that at least one absolute rendering intent is combined with at least one relative rendering intent during the transformation of the colour values of the intermediate colour space into the colour values of the second device-dependent colour space.

2. Method according to claim 1, wherein the at least one absolute rendering intent and the at least one relative rendering intent are each colourimetric rendering intents.

3. Method according to one of the preceding claims, wherein the intermediate colour space is a device-independent colour space.

4. Method according to any one of the preceding claims, wherein the colour values of the intermediate colour space correspond to the D50 2° normal observer.

5. Method according to one of the preceding claims, wherein the colour values of the first device-dependent colour space are transformed into the colour values of the intermediate colour space via at least one profile comprising at least one conversion table, and the colour values of the intermediate colour space are transformed into the colour values of the second device-dependent colour space via at least one profile comprising at least one conversion table, wherein the at least one absolute rendering intent and the at least one relative rendering intent use the same conversion table.

6. Method according to claim 5, wherein the at least one profile is at least one ICC profile.

7. Method according to any one of the preceding claims, wherein the device to which the first device-dependent colour space refers is a different device than the device to which the second device-dependent colour space refers.

8. Method according to one of the preceding claims, wherein the transformation of the colour values of the first device-dependent colour space into the colour values of the second device-dependent colour space is carried out during a printing process in which a digital print image is printed on a substrate by means of a printing press and wherein the first device-dependent colour space is the colour space of a screen and the second device-dependent colour space is the colour space of the printing press.

9. Method according to claim 8, wherein the printing method is selected from letterpress printing, planographic printing and gravure printing, in particular offset printing, inkjet printing, flexographic printing, screen printing and gravure printing.

10. Method according to one of claims 8 and 9, wherein no simulation of colour values of the unprinted substrate is carried out during the transformation of the colour values of the first device-dependent colour space into the colour values of the second device-dependent colour space.

11. Method according to claim 10, wherein the colour values of the unprinted substrate are paper white.

12. Method according to one of the preceding claims, wherein the transformation of the colour values of the first device-dependent colour space into the colour values of the second device-dependent colour space transforms the brightest point of the first device-dependent colour space to the brightest point of the second device-dependent colour space to an output without printing dots.

13. Method according to any one of the preceding claims, wherein the at least one relative rendering intent comprises black dot compensation.

14. A printing press with which a method according to any one of claims 1 to 13 is carried out.

15. Use of the method according to any one of claims 1 to 13 in a printing press.