Method for determining a match factor with regard to the color match of a produced printed product with specified target values for colors

Description

[0001] The present invention relates to a method for creating an objective evaluation standard for evaluating printed products.

[0002] The current state of the art for producing printed products essentially comprises the following: After a printed product has been "designed," technical parameters are defined. This includes determining the printing press on which the printed product will ultimately be printed. The press, in turn, specifies technical parameters. For example, it has a predetermined number of inking units, meaning it can process a certain number of colors. Furthermore, the inking units are arranged in a predetermined order, which also determines how colors can be printed consecutively on top of one another. Furthermore, the printing press to be used automatically specifies the printing process to be applied, which is also referred to as the "target printing process." For example, this could be an offset printing process. Furthermore, the substrate on which the printed product will be produced is specified.

[0003] The template for the printed product itself first reveals which final colors are to be achieved. In other words, a certain number of color appearances result in the printed product. This can, for example, be broken down into individual printable dots, and for each individual printable dot, the final color can be determined. This pixel-like processing is common.

[0004] These results, known as final colors, can be defined unambiguously from a technical perspective. For example, they have a unique spectrophotometrically measurable value.

[0005] Knowledge of the composition of colors is also state of the art. Any measurable color can be determined from the mixture of a few colors. "Mixing" in the context of the present invention does not necessarily mean creating a mixture or suspension in the sense of combining colors in specific proportions, but rather generally means creating a "final color" by overprinting specified colors in specified quantities and in a specific sequence. This is referred to as the so-called mixing intention or color intention.

[0006] Common printing ink systems, defined, for example, by the process colors CMYK (cyan, magenta, yellow, black), have now been developed to such an extent that measurable colors can be transferred into the respective color space. By overprinting the CMYK process colors in a specified order and at a specified percentage, a color impression can be created that is as close as possible to a given color.

[0007] It becomes obvious here that, on the one hand, the number and sequence of inking units in the printing press with which the printed product is to be produced and, on the other hand, the mixing intentions resulting from the color analysis, i.e. the specifications of the sequence and proportions of available colors to be printed on top of each other, must be coordinated with each other.

[0008] Another key aspect that influences the appearance and quality of a printed product is the substrate on which it is printed. The printing paper, its whiteness, any coloring, texture, absorbency, and the like can have a significant impact here.

[0009] All of the parameters mentioned above, such as the printing press, printing process, ink types, number of inks, printing paper, etc., are referred to as printing process parameters. Some of the printing process parameters commonly used in practice are standardized in international standards. For example, the color properties of the process inks and the coloring of the paper, along with other parameters for different paper classes, are specified for the offset printing process in ISO standard 12647-2 (Process Standard Offset Printing).

[0010] Once the printing process parameters for a planned print job have been defined, a data set representing the printed product is created, which can be processed by the designated printing press to produce the printed product. The press then prints the specified quantities of available colors, pixel by pixel, in the appropriate sequence, thus automatically producing the desired product.

[0011] A printing process can be measured by spectrophotometrically measuring the color mixtures printed on the sheet. This then determines the reflection spectrum of a color mixture and thus defines the color impression. In practice, often only a characteristic number of color mixtures are measured in order to limit the time required to a reasonable level. For the standardized printing conditions mentioned above, such measurements are often already available as state-of-the-art and have been published accordingly (e.g., FOGRA 51).

[0012] For a printing process with given printing process parameters, and thus also for a print data set generated for these printing conditions, spectrophotometrically measured target values, i.e. the reflection spectra of the individual color mixtures, are available. For better comprehensibility and clarity, the spectral target values are often converted into CIE L*a*b* values. This color system, defined by the CIE Commission, arranges different color impressions in a 3-dimensional space. The coordinate system used for this defines the individual color locations based on their brightness (L*), their position on a green-red axis (a*), and their position on a blue-yellow axis (b*). The coordinate system is constructed by definition so that the Euclidean distance in this space correlates with human perception of color differences.If the color coordinates of two colors are close together in this coordinate system, they are almost indistinguishable for our human perceptual system. This is regardless of which color mixture was used for printing. In other words, there can be different color mixtures that lead to an identical or indistinguishable color impression in the human perceptual system. The perceptible color difference is often expressed in the so-called delta-E 76 value according to the state of the art. This value describes the Euclidean distance in the CIE L*a*b* space. Current state-of-the-art developments have improved the calculation in the so-called delta-E 2000 formula. Color differences calculated using this formula correspond even more closely to the perceptual perception across all color ranges.For example, we perceive color differences in the yellow areas of the CIE-L*a*b* system less strongly than is described by the spatial distance there.

[0013] In summary, according to the state of the art, for a print data set defined for a known set of printing process parameters, a target value corresponding to the human color perception can be measured and thus determined for each color mixture or read from published measurements. This target value can be expressed as a color coordinate in a perceptual CIE-L*a*b* coordinate system.

[0014] Printing process parameters may need to be changed for a variety of reasons. For example, a printing process may need to be changed. For organizational reasons, it may be necessary or desirable to print at a different location, in a different country, on a different machine, using a different process, etc. This means that the data set representing the printed product must be changed, i.e., usually transformed. There are various methods for this, which generally involve color space transformations. This color space transformation describes new color mixtures for the modified printing process that are intended to correspond to the color impression of the original mixture under the original printing conditions. In practice, the question arises as to how close the printed product will come to the desired result after the transformation has been carried out.

[0015] In the following, the printing process with the modified printing process parameters is referred to as the target printing process, in the sense that this printing process represents the target of the transformation and to clearly distinguish it from the source printing process, i.e., the original printing process with its set of process parameters. Similarly, we refer to a source print data set and a target print data set.

[0016] In the prior art, it is known and common practice to verify the result of a transformation by comparing it with the original. Assuming that a template exists that corresponds to the requirements of the user, the target values for colors of the source print product to be printed in a defined color space can be determined through measurements. The source printing process parameters underlying the target result, i.e. the print data, are also known. The printing process parameters of the planned target printing process are also known. A transformation method is then selected, of which there are usually several different ones. This is followed by the transformation of the print data set for the print product to be printed for use in the target printing process using the specified transformation method.

[0017] An example of such a procedure is described in DE 10 2011 015 306 A1. In this process, test patches are printed using a reference printing process. This is a calibrated printing process. The corresponding test patch is then measured using a spectrophotometer to generate actual spectral data. From the resulting densitometric, colorimetric, or spectrally determined dot gains for the individual printing inks, corresponding dot gain correction curves are determined using interpolation methods. These curves can then be applied to the target printing process to achieve the target gray axis of the reference printing process. Thus, a target-actual comparison is performed for the purpose of correcting the target printing process. The transformation process is not evaluated.

[0018] If you now create a printout of the printed product together with measurable fields in the target printing process, you can spectrally measure these fields, which yields the actual values. You can then compare the target values with the actual values and thus make an evaluation.

[0019] If you repeat this process using a different transformation method, you can determine which transformation method is most suitable for the planned process.

[0020] State-of-the-art procedures are available for evaluating the transformation process. The transformation process is often evaluated using visual matching. This involves visually comparing the prints from a source dataset with the transformed and also printed versions from the target printing process and evaluating them by experts. Known and deemed suitable motifs are used. The "Roman 16" image dataset provides a set of such motifs. A characteristic of the structure of this set of motifs is that individual color ranges dominate in the individual images. This makes it possible to separately evaluate the quality of the transformation process in individual color ranges, as well as the quality in neutral gray tones in the medium brightness range, independent of dark and saturated tones.The set of motifs can also be expanded to include motifs reflecting the use of spot colors, such as those frequently used in packaging printing. Here, too, the motifs can be structured in such a way that individual quality aspects can be evaluated separately. Ultimately, however, such an evaluation is time-consuming, often requires a large number of experts, and is subject to highly subjective observer criteria.

[0021] As explained above, the target values for a source print data set can be determined spectrophotometrically and the corresponding color locations defined in the perception-specific CIE-L*a*b*. Actual values for the transformed target print data set can be determined in the same way. According to the state of the art, the difference in perception can now be calculated for each individual color mixture using a delta-E 2000 value. The visibility threshold is generally specified with values below 1, i.e. if only values smaller than 1 can be determined for the color mixtures in comparison, the transformation process has a high quality such that the target print data set does not visually differ from the source data set.However, this simple method, a method for statistically evaluating all color differences between all color mixtures, reaches its limits when the target printing process, due to its physical properties, is unable to reproduce all color locations of the source print data set. The overall statistics can then no longer be reduced to values below the visibility threshold even with a theoretically optimal method, and thus individual or multiple larger differences distort the statistics and thus comparability and evaluability.

[0022] One example of this is printing on a significantly modified target printing substrate. The source data set was prepared for bright white art paper. It is now printed on brown cardboard. As a result, not all of the light color mixtures of the source data set can be achieved in the target printing process. Even the unprinted area already exhibits high visible differences that cannot be resolved even with a transformation process. State-of-the-art technology allows for a media-relative comparison in this case. The target values to be achieved are transformed in such a way that the differences in the unprinted target printing substrate are eliminated. The quality of the transformation process can then be evaluated using statistical methods. This evaluation method is described, for example, in the Fogra Process Standard for Digital Printing (PSD).However, even the state-of-the-art media-relative evaluation method exhibits significant weaknesses if the achievable color gamut of the target printing process is even more severely restricted. The color gamut is defined as the set of all possible color locations that can be achieved using a set of printing process parameters. An example of such a restriction is the darker and / or colored printing substrate mentioned above. In this case, the color gamut in this area is already severely limited. If, for example, the process colors can no longer be applied with the same intensity or overprinted in the same quantity in the target printing process, the range of all representable colors shrinks further. The narrower the color gamut of the target printing process, the less important the purely statistical consideration of color differences after transformation becomes.For example, it is possible to map many color locations from a source print dataset to a single color location in the target print dataset if the target print dataset does not contain these color locations in its color gamut. Statistically speaking, such a transformation can have a smaller overall or average error and thus rate a transformation better in terms of its quality than an alternative transformation, which in this case preserves the distinguishability of the color locations in the source print dataset but leads to a higher error when purely considering the deviation. An expert who performs a visual assessment of the transformation as described above would rate the alternative transformation as better in this case. In this case, the objective methods known according to the current state of the art are inferior in their assessment to an expert.

[0023] Based on the state of the art described above, there is a significant need to be able to determine the most measurable and objective evaluation of printed products after changes to the data set representing the printed product, in order to also be able to specifically assess the deviation from the specification, while largely excluding individual influences and subjective evaluation impressions. The method must also meaningfully include target printing process parameters in its evaluation statement that have a significantly narrower color gamut compared to the source print data set.

[0024] The object of the present invention is to provide a method for determining a matching factor with regard to the color matching of a produced printed product with predetermined target values for colors.

[0025] To solve this problem, a method having the features of patent claim 1 is proposed. Further advantages and features of the invention emerge from the subclaims.

[0026] The starting point is a uniquely identified printed product, which has a clearly defined final color appearance. Typically, the color appearance for each individual pixel is specified in a perception-specific and printing-independent color system. One such suitable color system is the CIEL*a*b* system. A print data set exists for this uniquely identified printed product.

[0027] It should be noted here that common printing processes are often standardized with regard to their printing process parameters, such as paper, inks, color sequences, and the like. Idealized values for color appearance are often published for these standardized printing processes. For example, the FOGRA51 data set for offset printing on Class 1 art paper of ISO standard 12647-2.

[0028] Typically, the target printing process is then determined and a transformation process selected. These processes are implemented by people involved in printing. The target printing process is usually largely predetermined due to availability, organizational, or design considerations. Various transformation processes are then available to transform the existing print data into the print data required for the planned print using the original printing process. The print is then executed, and the result is measured to determine the actual values. Like the target values, the actual values are also measured spectrophotometrically and converted to the CIEL*a*b* color system.

[0029] According to the invention, a color value axis is now created. For this purpose, an axis is defined from a unique color value (starting point) to a second unique color value (end point), while simultaneously specifying the support points selected between the two color values. This is also done by a person involved in printing using a computer with the relevant data available. This selection can also be made in conjunction with the selection of the uniquely identified printed product.

[0030] From the set of target values for the uniquely identified printed product, a computer automatically determines the target values for the start point, the end point, and the support points. The computer then extracts the actual values for the start point, the end point, and the support points from the set of measured values for the product printed and measured using the target printing process. The distance between the target value and the actual value for each of the start points, end points, and support points can now be automatically determined, and a distance value representing this distance can be calculated. In this initial evaluation, the distance is advantageously determined using the delta-E 2000 formula, as this most closely approximates human perception.

[0031] A match factor can now be calculated and output as a statistical average from the distance values. According to the invention, percentages are used here. This allows a single distance value to be expressed as a percentage. If the target value and actual value completely agree, the match is 100%. Accordingly, deviations can be expressed as a percentage. The distances, expressed as percentages, can now be averaged along the color value axis to produce a single value. This represents the color match of the printed product with the specified target values for colors.

[0032] The next step in the invention is to perform a coordinate transformation of the measured values. This coordinate transformation maps the respective starting points of the color axes to each other, as well as the endpoints of the color axes. The intervening support points are modified using the same conversion rule. The Euclidean distance between the points can now be determined for the support points in the thus transformed coordinates. The distances can be converted into percentage notation and averaged in the same way. This results in a further agreement factor. This evaluates the extent to which color differences between the support points along the color axis in the target printed product were retained by the color space transformation being evaluated.If the two color axis match factors are averaged, a final value is obtained that describes both the color match of the resulting printed product with specified target values and the preservation of the original color differences between the color axis support points. This aspect, which evaluates the so-called color modulation of the printed product, is particularly crucial for an objective evaluation of the transformation process and the resulting printed product in the target printing process. Since a specific transformation process was used, it is now possible to make an objective comparison between different transformation processes.If the same procedure using a different transformation method results in a different matching factor, this provides an objective measure of the suitability of the transformation method for transforming print data in relation to the specific target printing process.

[0033] Such findings can be stored and retrieved in databases, from which the most suitable transformation methods can be selected for future printing processes.

[0034] According to a further development of the invention, additional color axes can be created and evaluated using the method described above. This makes it possible to include all relevant color ranges of the printed product in the evaluation by creating separate axes for each color range, with suitable start and end points and the corresponding support points. For the example of the "Roman 16" motifs, the dominant colors in each motif can be placed on the axes as already specified there and then evaluated using the method proposed in the invention.

[0035] Furthermore, according to a further development of the invention, the matching factors of all color axes can be converted into a single matching factor. According to a further development of the invention, it is advantageous to assign different weights to different color ranges. Similarly, the weight assigned to the two values for each color axis when determining the matching factor can be changed depending on the color range. This allows, for example, greater weighting to be given to modulation on a gray color axis, while greater weighting to color matching in the red range is given to color matching. The scheme used to distribute the weights across the different color axes allows for an objective and clearly defined evaluation of the printed product.

[0036] Instead of an arbitrary review of target / actual comparisons, a targeted determination of a value is now carried out by specifying a specific verification quantity.

[0037] As already mentioned, the standard templates mentioned and described above are suitable for printing, for example, the "Roman 16" motifs based on process colors. These templates can now also be used according to the method according to the invention to determine a matching factor.

[0038] According to a particularly advantageous development of the invention, templates can be designed that also incorporate spot colors into the design of the source printed product. For example, the CMYK process colors can be supplemented by one or more defined additional colors. For example, Pantone color ranges are known for which clear target values are available. Accordingly, templates for printed products can be developed which, when applied to the determination method according to the invention, allow concrete statements to be made about the suitability of a particular transformation method for specific printing processes. For example, when using certain brown tones in packaging printing on a certain machine type, the application of only certain transformation methods can be determined.The ability to evaluate the transformation of spot colors from the source printing process to the target printing process is particularly important when it comes to digital printing. Digital printing processes very often do not use spot colors, but rather an expanded, fixed set of colors. For example, the process colors CYMK are supplemented by orange and green. This means that the spot colors are necessarily changed by the transformation process. In this respect, evaluation using the inventive match factor is very important, since here too the quality of the printed product created in the target printing process depends not only on the color match of the spot color, but its color gradations also represent a key quality criterion.

[0039] The invention proposes a method that enables the determination of a matching factor independent of individual or subjective influences. This factor provides information about the color match of a printed product to a clearly identified and color-defined original. Since the print data of the original were transformed for output of the printed product in the target printing process using a data transformation method, the matching factor indicates a value for the suitability of the transformation method for the specific printing process.

[0040] This will enable printing processes to be optimized quickly and specifically in the future by using the appropriate transformation method. This can take into account the colors used, base systems and spot colors, overprint specifications and mixing intentions, the printing substrates and their colors, the printing process, the printing press, and so on. Thus, if any printed product is available in the future, reproduced using the print data related to a specific printing process, the appropriate transformation method can be identified to produce this printed product using a different printing process, achieving the greatest possible color identity or color fidelity.

[0041] The method according to the invention is carried out automatically after the specifications have been recorded and automatically generates a final value in the form of the conformity factor. It solves the technical problem of obtaining a reliable evaluation standard, free from individual and subjective influences, in order to be able to make well-founded decisions about the applicable procedures.

[0042] Further advantages and features of the invention will become apparent from the following description based on the figures. Figure 1 shows an exemplary representation of a printed product comprising CMYK process colors and spot colors, as well as the patch fields as a basis for target values; Figure 2 shows a schematic representation of a three-dimensional CIEL*a*b* coordinate system with color axes; Figure 3 shows a schematic representation of a standardization for determining distances; and Figure 4 shows an exemplary representation of a result of determining a match factor.

[0043] As already explained, the industry standard formula used to determine color differences is the DeltaE formula. This formula describes the Euclidean distance in the so-called CIEL*a*b* color space, or the improved formula DeltaE 2000. If two points in this space are so close together that their distance is less than one (DeltaE < 1), they are no longer visually distinguishable. If you now reproduce a motif whose pixels represent the reference data set expressed as Lab values on a different output system (different printer, different format, different inks, etc.), a test data set consisting of CIEL*a*b* values is created. If every pixel is reproduced down to DeltaE < 1, no visual difference can be perceived between the original and the reproduction, and you get optimal reproduction.

[0044] However, the output system usually defines physical conditions that make such optimal reproduction impossible. The best example of this is a substrate that has a different color than the reference. For example, if you print an image on yellow newsprint, an exact reproduction of the CIEL*a*b* value, i.e., a DeltaE < 1, is impossible for many pixels. Thus, while the DeltaE formula can indicate that identical reproduction is not possible, it does not provide a measure of optimal reproduction within the physical conditions on a specific output system. Such a formula was developed by GMG.

[0045] It is necessary to determine the extent to which the spatial relationships between the pixels of the reference data set are reproduced in the test data set. For this purpose, selected color values (arranged as measurement fields next to the images) from the reference data set, as well as the corresponding values from the test data set, are subjected to a transformation into a new coordinate system. Two points are selected that, in the new coordinate system, point to the origin or to the point (x = 100, y = 0, z= 0). All other points undergo a corresponding transformation into this new three-dimensional space. In this space, the relative distances between the reference and test data sets can now be both measured and visually represented. The metric is designed so that if all points lie on top of each other after the transformation, the best possible test data set has been achieved, since the relative spatial relationships have been preserved. Deviations from this optimal representation can now in turn be expressed as Euclidean distances in this new space between the individual reference points and their reproduction (the test data set).

[0046] Figure 1shows sample templates created from combinations of CMYK process colors and clearly identified spot colors. The measurement fields of the individual representations represent the color ranges relevant to this image and thus the axes formed. These represent the target values for the respective colors. For each of the images, there is a unique print data set, which is transformed into the desired target printing process using a transformation method as required. The transformed data is then printed in the target printing process, and the measurement fields are measured spectrophotometrically accordingly. This results in the actual values.

[0047] Figure 2shows a three-dimensional CIEL*a*b*-based coordinate system for representing the color axes for different color ranges. A starting point, an end point, and intervening support points are defined for each color axes. The target values and the actual values of each color axis are positioned in this coordinate system. As described above, the distances can be specified as Euclidean distances in this space between the individual reference points and their reproduction.

[0048] This principle is in Figure 3 On the left, the reference data set is shown in blue in the CIEL*a*b* color space. This must be reproduced on an output system that, due to physical limitations, has a smaller color space, thus forcing a visible color difference (DeltaE >> 1). The test data set is shown in red.

[0049] On the right side you can see how by normalizing the top left point to the coordinate origin and the bottom right point to the coordinates x=100, y=0, you can achieve a representation that shows how well the relative spatial relationships are preserved. Figure 4shows the result of the procedure according to the invention. For the individual color areas based on the prints of clearly identified originals, the distances can be specified as percentage values. Different values result depending on whether the values are evaluated absolutely as a distance across the entire color axis or relatively after standardization. The individual values of the color areas can now be summarized to form an overall value, which in the case of the present example was 86%. This value represents the match factor that results from applying the selected transformation method to produce a printed product in a specific target printing process. If the same procedure is now repeated using a different transformation method, clear and comparable end values result, which enable an objective selection of the appropriate method.

Claims

1. Method for determining a conformity factor with regard to the colour conformity of a printed product with specified target values for colours, comprising the following steps: a) Determining the target values for colours of the printed product to be printed in a defined colour space in a source printing process, b) identifying the print data set specified for a target printing process based on printing process parameters that have been changed with respect to the source printing process, c) transforming the print data set for the printed product to be printed for use in the target printing process using a specified transformation method, d) Generating a printout of the printed product together with measurement fields of this print in the target printing process, e) Spectrally measuring the measurement fields of the printout according to d) to record the actual values, f) Comparing the target values with the actual values, characterised by the successive steps: g) Defining a colour value axis from a unique colour value as the starting point to a second unique colour value as the end point, while simultaneously defining the reference points between the two colour values, h) Determining the target values for the starting point, the end point and the reference points, i) Determining the actual values for the starting point, the end point and the reference points, j) Determining a distance value representing the distance between the target value and the actual value for each of the start points, end points and support points, k) Calculating and outputting a conformity factor in the form of percentage values as a statistical mean value from the distance values.

2. Method according to claim 1, characterised in that that the start and end points of both the target values and the actual values are brought into alignment by standardisation, resulting in a transformation rule, and the reference points of both the target values and the actual values are subjected to the same transformation, whereby the resulting distances of the actual values from the target values of the reference points are used as distance values to determine the conformity factor.

3. Method according to at least one of the preceding claims, characterised in that a plurality of colour axes are defined for a printed product and processed according to steps h) to k).

4. Method according to claim 3, characterised in that a resulting factor is formed statistically from a plurality of conformity factors.

5. Method according to at least one of the preceding claims, characterised in that a printed product is selected which comprises the process colours CMYK.

6. Method according to at least one of the preceding claims, characterised in that a printed product is selected which has a spectrum of a special colour.

7. Method according to at least one of the preceding claims, characterised in that the determined matching factors are stored in a database in a retrievable form together with the information about the printing process, the number of colours and colour selection, and the respective data processing method for preparing the print values.

8. Method according to claim 7, characterised in that, for a planned printing project, a suitable data processing method for preparing the print values is selected by evaluating the database information.