WO2007137624A1

WO2007137624A1 - Ad-hoc color gamut representation

Info

Publication number: WO2007137624A1
Application number: PCT/EP2006/062823
Authority: WO
Inventors: Oscar Martinez; Santiago Felez
Original assignee: Hewlett-Packard Development Company, L.P.
Priority date: 2006-06-01
Filing date: 2006-06-01
Publication date: 2007-12-06

Abstract

A printing system includes a printer arranged to render color images on print media and having a color gamut; a color measurement device embedded in the printer and arranged to measure colors, printed by the printer on a print medium, in a device-independent color space, and a color gamut constructor arranged to form, in an ad-hoc manner, a model representing the printer's color gamut with its boundary, based on a color measurement by the color measurement device of a set of color patches printed by the printer. The color gamut constructor is comprised in the printer or in a printer-driving program module arranged to run in a computer connected to the printer, or in both.

Description

AD-HOC COLOR GAMUT REPRESENTATION

FIELD OF THE INVENTION

The present invention relates generally to color imaging, and for example, to printing systems and methods for forming a representation of a printer's color gamut.

BACKGROUND OF THE INVENTION

Color management is becoming more and more important since nowadays not only professional designers and illustrators but also average consumers wish to reproduce color as faithfully as possible by means of their electronic color output devices. This development is, among other things, the consequence of a rapidly increasing number of users taking pictures with digital cameras and printing the pictures, e.g. with their desktop color inkjet printers. One crucial point in terms of color management is that each color recording or reproducing device has its own device-dependent color space by reference to which it records or reproduces colors. For example, two digital cameras of different manufacturers taking the same picture under the same lighting conditions will store different RGB-values in their memory due to the differences between their photo-sensors, lens-systems and color processing firmware. Therefore, in order to be able to compare RGB-values of different color input devices, the colors are integrated into one common color space, which is usually the CIE (Commission Internationale de IΕclairage)-LAB-color space or the CIE-XYZ color space. The LAB-color space has the property that it indicates the color values as perceived by the receptors in the human eye and is therefore also referred to as "psychometric color space". Furthermore, the LAB-color space is "perceptually uniform" which means that changing any attribute of a color by the same increment will produce the same degree of visual change (see for example: B. Fraser et al.,"Real World Color Management", Peachpit Press, 2003, p. 69-72). Another example would be two color inkjet printers of two different manufacturers operating in a CMYK-color space having the primaries cyan (C), magenta (M), yellow (Y) and black (K). If the same CMYK-values are sent to the different color printers, perceptually different colors will appear on the print medium with regard to the LAB- color space. (Theoretically, the color black can be composed by superimposing cyan, magenta and yellow, each of C, M and Y at 100%; apart from the fact that this approach would consume an excess of ink, print media have certain ink limits not to be exceeded, and the colorants used in color inkjet printers do not yield a genuine black because of their chemical properties. Therefore, black (K) is added as a primary color.) The LAB-values of color patches printed on a print medium can be measured e.g. with a spectrophotometer which yields the LAB-color values of the color patches.

If LAB-values of some color patches are measured, the printer-related CMYK- values of which are known, LAB-values can be determined for all possible CMYK- values by means of a mathematical transformation. To this end, color values are transformed from the device-dependent color values (e.g. RGB-color space of a scanner or digital camera) into the device-independent color values of the LAB-color space. This transformation may be performed by means of a profile which represents a mapping from the device-dependent color space of a color device into the LAB- color space. To this end, a "neutral observer" is applied which is able to measure a color in the LAB-color space. This neutral observer is typically a colorimeter which uses filters that mimic the neutral observer's response or a spectrophotometer which measures the wavelengths of the reflected light of color patches and calculates the corresponding LAB-color values. Typically, a spectrophotometer is an external device, often a hand-held device (such as Gretag Macbeth EyeOne Pro), which measures the device-independent LAB-color values of a set of color patches. A set of color patches printed by a color output device on the basis of known device- dependent color values is also referred to as a "target". If the color patches have been produced, e.g. by an inkjet printer having its own CMYK-color space, whereby a point in the CMYK-color space represents a corresponding mixture of the four different inks cyan, magenta, yellow and black, then the CMYK-color points associated with the color patches can be assigned to the measured LAB-values of the patches to obtain a profile. In this context, the color values provided to the color output device are also referred to as "stimulus" and the color patches printed are referred to as "response" of the color output device to the color values. Thereby, a mapping is defined which maps the device-dependent color values (e.g. CMYK values to the LAB-values of the color patches measured by a spectrophotometer). This mapping is often represented in the form of a matrix or a lookup-table (LUT), whereby a lookup-table representing a mapping from a device-dependent color space into a device-independent color space is usually referred to as "AtoB"- mapping, and a mapping from the device-independent color space into the device- dependent color space is usually referred to as "BtoA"-mapping. Often, a target has more color patches than a lookup-table has entries. Then, an interpolation of the measured color values is performed to reduce their number to that of the entries.

A color gamut of a color reproducing device, which is the set of all colors reproducible by the device, can be visualized in the LAB-color space. Thereby, the color gamut of a color reproducing device only constitutes a subset of the entire LAB- color space, which is the color space of all colors visible for a human being. Generally, the question arises how to represent an image containing colors outside the color gamut of a color reproducing device, i.e. colors that the color reproducing device cannot reproduce, or out-of-gamut colors. To this end, "rendering intents" have been introduced to map those colors which are outside the gamut (represented in the device-independent color space) of the color reproducing device to colors which are inside the color gamut of the corresponding color reproducing device. Incidentally, a "rendering intent" - as connoted by the word "intent" - may be a subjective way of mapping out-of-gamut colors to reproducible colors. In other words, rendering intents are auxiliary means to map colors which cannot be reproduced otherwise. As there is no perfect way of mapping these colors (since this mapping always introduces color changes), a user can opt for a rendering intent. Four different general ways of rendering intents are used and normally incorporated in different lookup-tables (see for example: B. Fraser et al., p. 88-92).

In US 6,809,855 B2 an "improved and lower cost color spectrophotometer" is described, which is integrated in a color printer for on-line continuous color correction purposes.

US 2002/0080373 A1 describes a proofing printer with an embedded color measurement device for emulating a high volume output device. To this end, a target ("test image") printed by the proofing printer is measured first to calibrate the proofing printer ("color calibration adjustments"), and then a target printed by the high volume output device is measured to modify the proofing printer's calibration so that it emulates the high volume output device ("color management adjustments"). In US 2002/0080373 A1 , the color gamut of the proofing printer or the high volume output is not explicitly determined, but can be seen as implicitly included in the targets by the absence of certain colors (since a target does not show any colors that cannot be printed by the device considered), and it is therefore implicitly measured when the colors of a target are measured. Thus, the mentioning in US 2002/0080373 A1 that the color calibration and management adjustments are based on "objective measurements of the color gamut and tone characteristics of the test images" (paragraphs [0010] and [0012]), and that the "color gamut and tone characteristics of the code values .... match the color gamut and tone characteristics of the output of the high volume output device that has been profiled" (paragraph [0055]), only refers to such an implicit gamut measurement.

US 2002/0140701 A1 describes a method and system for constructing and visualizing a model of a color gamut. A set of color data points is put in an additive color space, such as the CIE-XYZ space, and a convex hull is constructed from the color data points. Subsequently, the convex hull is transferred from the additive color space into a corresponding solid object in a psychometric color space, such as the CIE-LAB color space.

SUMMARY OF THE INVENTION

According to one aspect, a printing system is provided. It comprises a printer, a color measurement device, and a color gamut constructor. The printer is arranged to render color images on print media and has a color gamut. The color measurement device is embedded in the printer and arranged to measure colors, printed by the printer on a print medium, in a device-independent color space. The color gamut constructor is arranged to form, in an ad-hoc manner, a model representing the printer's color gamut with its boundary, based on a color measurement by the embedded color measurement device of a set of color patches printed by the printer. The color gamut constructor is comprised in the printer or in a printer-driving program module arranged to run in a computer connected to the printer, or in both.

According to another aspect, a method is provided of forming a color gamut model. The method includes printing a set of color patches with a printer on a print medium. The colors of the set of color patches are measured in a device- independent color space by a color measurement device embedded in the printer. A model representing the printer's color gamut with its boundary is formed in an ad-hoc manner by a color gamut constructor, based on the color measurement of the set of color patches. The color gamut constructor is comprised in the printer or in a printer- driving program module arranged to run in a computer connected to the printer, or in both.

Other features are inherent in the products and methods disclosed or will become apparent to those skilled in the art from the following detailed description of embodiments and its accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will now be described, by way of example, and with reference to the accompanying drawings, in which

Fig. 1 illustrates a printing system with a printer having an embedded color measurement device, and with a gamut constructor, according to embodiments of the invention; Fig. 2 illustrates how a color gamut construction is invoked, according to embodiments of the invention;

Fig. 3 shows an exemplary user interface, according to embodiments of the invention;

Fig. 4 illustrates a color-catalogue customizer, as an example of an application invoking an ad-hoc gamut determination according to embodiments of the invention;

Fig. 5 illustrates a comparison of gamut models of different printers or media/colorant/print mode combinations (Fig. 5a), and of an image gamut and a printer gamut (Fig. 5b), according to embodiments of the invention;

Fig. 6 illustrates "proofing" using embodiments of the invention; Fig. 7 illustrates a comparison of gamut models of proofing printer, according to embodiments of the invention, for different media/colorant/print mode combinations and of a printing press;

Fig. 8 is a flow chart illustrating a gamut model comparison, according to embodiments of the invention, Fig. 9 is a flow chart illustrating an automatic gamut model determination and comparison for different print media, according to embodiments of the invention;

Fig. 10 is an exemplary color-stability diagram;

Fig. 11 is a flow chart illustrating an additional color-stability functionality of some of the embodiments of the invention. The drawings and the description of the drawings are of embodiments of the invention and not of the invention itself.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Fig. 1 illustrates a printing system with a printer having an embedded color- measurement device, and with a gamut constructor. However, before proceeding further with the description of Fig. 1 , a few items will be discussed.

A gamut is the entire range of colors reproducible by a particular printing device (e.g. a color inkjet printer with a certain medium like glossy paper). "Out-of- gamut" colors are colors that are not reproducible. The gamut surface is the surface of the volume which includes all reproducible colors, in a color space. The gamut surface thus separates the color-space region of the reproducible colors from that of the out-of-gamut colors. In some of the embodiments, a printer is provided which is arranged to render color images on print media and which has a color gamut. The printer includes an embedded color measurement device which is arranged to measure colors which are printed by the printer on a print medium in a device-independent color space. The printer further includes a color gamut constructor which is arranged to form, in an ad- hoc manner, a model representing the printer's color gamut with its boundary, based on a color measurement by the embedded color measurement device of a set of color patches printed by the printer.

In some of the embodiments, the model which represents the printer's color gamut with its boundary is a parameterization of the gamut, for example a vector- graphics parameterization, such as a wireframe model (which models the gamut by surface points connected by lines), a surface model (which models the gamut boundary), or a volume model (which models the color-space volume occupied by the in-gamut-colors). Since the gamut surface encloses the gamut volume, and the boundary of the gamut volume is the gamut surface, a model representing the surface (e.g. wireframe or surface model) is also a representation of the gamut volume, and a model representing the gamut surface is also a representation of the gamut volume. A picture of the parameterized gamut can be displayed or printed on a screen or print medium, and not only the gamut parameterization, but also its display or print-out is called herein a "model which represents the printer's color gamut with its gamut boundary". In alternative embodiments, the gamut is not parameterized, but the measured colors of a target (set of color patches) printed by the printer are modeled and displayed or printed as a "cloud of points" in a color space on a screen or print medium, and when looking at the displayed cloud of points, the user will combine the points and recognize the gamut with its surface. Thus, display or printing of the color points in a color-space representation is also regarded herein as a "model which represents the printer's color gamut with its gamut boundary".

The gamut is typically a complex structure depending on various physical parameters so that it may only be determined with a certain accuracy which is increased with the number of measured colors. Therefore, the gamut model typically approximates the printer's real gamut. In a parameterization, often simplifying assumptions are made (e.g. that the surface is composed of flat segments), which is another source of approximation. In some of the embodiments, the gamut model is constructed as a convex hull from the measured color points. One exemplary suitable program for determining a convex hull from color points is the "Qhull" program, which is, e.g. available from the University of Minnesota. The determination of a convex hull involves the transformation of a cloud of individual data points into a solid object. In some of the embodiments, the construction of the convex hull is performed in an additive color space, such as CIE XYZ. If the measured colors are initially expressed in a non- additive color space, such as CIE LAB, then a color transformation from CIE LAB to the additive color space (e.g. CIE XYZ) is performed on the color data points, before the convex-hull construction is carried out. If the color data points are measured in an additive color space, then no such color transformation is performed. After the gamut model construction, since gamut models are, in some of the embodiments, represented in a psychometric color space, such as CIE LAB, a transformation from the additive color space to the psychometric color space is applied to the convex hull. This transforms the convex hull into a corresponding solid object in the psychometric color space, which is then, for example, displayed on a display device, or quantitatively analyzed in various ways. Further details of a convex-hull construction can be taken, for example, from US 2002/0140704 A1.

In some of the embodiments, the color gamut constructor is comprised in the printer. For example, the printer has a processor which is programmed, e.g. by firmware, to function as a color gamut constructor.

In other embodiments, the color gamut constructor is a part of a printer-driving program module arranged to run in a computer connected to the printer. The expression "arranged to run" includes a representation of the gamut constructor program which is executable, not yet installed on a computer, and, for example, is stored on a data carrier (e.g. a CD-ROM or DVD usually sold together with a printer) or is in the form of a signal propagated over a network (e.g. a file downloadable via the Internet). The expression "arranged to run" also includes a representation of the gamut constructor program which is installed on a computer connected to the printer, and is ready to run upon invocation.

In further embodiments, the color gamut constructor is comprised partly in the printer, and partly in the printer-driving program module. For example, a first parameterization of the gamut is performed by the printer's processor and sent to the connected computer, and further parameterization and/or display tasks are then performed by the printer-driving program module running in the computer.

Initially, programs to drive printers only processed "printer-independent" printing information to "printer-dependent" printing information usable by the corresponding printer to produce printouts. Early personal computer printers could, for the most part, only print characters. Functions like the positioning and definition of margins and fonts were indicated by special codes sent by the personal computer.

The codes for printers from different manufacturers were incompatible with each other, and each application program had a different set of "printer drivers" with those codes embedded. As software became more capable - word processors, presentation graphics, and page layout programs in particular - this technique became impractical. Different application programs had a different set of drivers. If a user bought a new printer, he had to get new drivers from each of his software vendors. The software vendors had to create and maintain their printer driver libraries. Modern operating systems, like Windows©, simplified this problem, because they provided a printer-independent interface between application programs and the printers. In order to reduce the processing and network load, some printer drivers use what is called a page description language (e.g. PostScript© or PCL©) to communicate what is required to the printer. Notwithstanding the fact that many printer manufacturers have standardised on a page description language, printer drivers specific to the respective printer are still needed. This is mainly because the printer drivers allow the control options and features in the printers, and the controls for them are not standardised. Most printer drivers are based on a core printer driver, developed by the provider of the operating system, which handles things like rendering fonts, choosing the port the printer is connected to, and otherwise performing the tasks needed to send a page to the printer. The printer manufacturer builds tables that tell the core driver how to accomplish the necessary functions, and adds functions, for example related to colour matching (Barry Press et al.: PC Upgrade and Repair Bible, 1999, third edition, pages 746-749). All these features form a printer-driving program module. As the color gamut constructor is (partly or completely) comprised in the printer and/or the printer-driving program module, in some of the embodiments, this enables the user to invoke the process of forming a gamut model (printing a target, measuring its colors, constructing the gamut model for a certain print- medium/colorant/print-mode combination) from the user interface also used for normal printing. In further embodiments, the gamut-model-forming functionality is accessible through the printer-driving program module on the operating-system level, which enables application programs that require a gamut model as input to invoke the gamut-model-forming functionality and get the gamut model returned.

For example, graphics designers often use catalogues of discrete colors which cover the whole color space in a structured and simple-to-handle manner (such as the Pantone© color catalogue). However, if such a catalogue is to be used on a printer, a (reduced) printer's gamut often does not enable all the colors of the original catalogue to be printed. Suitable applications map the catalogue to a "customized" catalogue the colors of which are in gamut, while trying to preserve the catalogue's structure. The actual mapping to be performed depends on the printer's gamut (for the print-medium/colorant/print-mode combination considered), so that such an application performs a gamut-depending color mapping. In the embodiments mentioned, the gamut-model-forming functionality is accessible through the printer- driving program module by the catalogue-mapping application. When a catalogue is to be customized to the connected printer (for a certain print-medium/colorant/print- mode combination), the application can invoke the gamut-model-forming functionality (either automatically or by a user from an interface provided by the application), and the printer-driving program module will form the gamut model (by initiating printing a target, measuring its colors, constructing the gamut model for the print- medium/colorant/print-mode combination) and return data representing the gamut model obtained (e.g. a parameterization of the gamut model). The calling application will use the returned gamut model as an input, e.g. to perform the mapping of the original color catalogue to an in-gamut catalogue. The ability to form the color gamut in an ad-hoc manner is due to the fact the color measurement device is embedded in the printer and that the gamut constructor is also functionally embedded, by being comprised in the printer and/or being a part of a printer-driving program module arranged to run in a computer connected to the printer. This enables a target to be printed, and, ad hoc and in the same printer, its colors to be measured, either immediately or after an ink-drying delay (of, e.g. one hour), and the gamut model to be formed.

Generally, the color gamut depends on the type of print medium (e.g. glossy or coated paper) and the type and number of colorants (e.g. opaque or transparent, three, four or more inks), as well as the condition of the colorants (e.g. age or temperature) used for a printout. Another factor which influences the color gamut of a printer is the print-mode used. For example, a "draft print" in which printing is faster and coarser will generally result in a different gamut to a "quality print-mode" which prints slower and finer. The "print mode" may also include other print-process parameter, e.g. whether a print is made line-by-line, or in an interwoven mode. This may influence the extent to which overprint occurs and, consequently, the color gamut. Therefore, in some of the embodiments, the printer's color gamut depends on at least one of: the print medium used, a colorant used, and a print-mode used. Hence, a gamut model formed will be specific for the given print- medium/colorant/print-mode combination used for printing the target. When only a small number of inks are used (e.g. four in a CMYK printer), colors other than the inks are usually produced by halftoning techniques. To produce a certain color in a viewer's perception by halftoning, dots of different inks are normally printed side-by-side, but to a certain extent also on top of each other. This extent will generally differ between different types of halftoning (e.g. whether and what regular screen, or random-dot arrangement is used), so that the gamut will also depend on the type of halftoning used. In some embodiments, the printer implements different halftoning types that can be selected by a user. Such different halftoning types are also subsumed under the expression "print mode" herein.

Incidentally, a color gamut of a printer is not completely static, but will rather depend on the condition of the printer, e.g. whether its mechanics are new or worn out. Thus, the ad-hoc color-gamut model also represents a "snapshot" of the current condition of the printer.

The term "print medium" as used herein refers to any types of paper, such as glossy, semi-gloss or coated paper, different types of transparencies, cardboard, canvas, and any other substance on which a printer can print.

The term "colorant" refers to dye ink, pigmented ink, toner, color-coated film for thermal printers or any other substance which can be applied to a print medium.

The term "printer" as used herein refers to any sorts of inkjet printers, printing presses, color laser-jet printers, thermal printers or any other devices which are able to print color on a print medium.

In some of the embodiments, a "color measurement device" is a colorimeter or spectrophotometer. Colorimeters directly measure colorimetric values by suitable color filters that mimic the human cone response, and produce numerical results in a color space (e.g. CIE LAB). Spectrophotometers measure the entire spectrum, but process the data representing the measured spectrum such that the response to the spectrum by the cones in our eyes is simulated, and again output numerical results in a color space (see for example: B. Fraser et al., p. 43-44).

The set of color patches whose color values (in the color space of the printer) are known is referred to, especially in the context of color management systems, as a

"color target". The targets are created by presenting different color codes that span the printer's color print range, i.e. combinations of different amounts of the different colorants to the printer in an array that covers the range of possible combinations of the colorants. Thereby, the printer is caused to print a variety of different colors defined by the color codes. The inks are, for example, CMYK; thus, the color codes are expressed in the (device-dependent) CMYK color space. However, there are also printers using more than four inks, which additionally provide a light cyan and a light magenta, thereby operating in a CMYKcm color space. Other printers are equipped with inks of the four subtractive primary colors cyan, magenta, yellow, black and further provide an orange and green ink. These printers refer to a CMYKOG color space, whereas still other printers even operate in a CMYKcmOG color space. The presentation of the set of color codes (i.e. all the different combinations of CMYK-inks) to the printer is therefore referred to as a "stimulus". A response is obtained by measuring the colors of the different color patches with a spectrophotometer or colorimeter in a device-independent color space. In the usual calibration procedure such a measurement is used to construct a "profile" by assigning the measured LAB-color values to the known CMYK values (e.g. in US 2002/0080373 A1 ). Such a profile indicates what actual colors will result from a given set of CMYK-values, and the reversed profile indicates the inverse mapping from actual colors to CMYK-values and is normally used to tell a printer what to print. The gamut of the printer - which is of interest here - is inherent in the target and the color values measured, but only implicitly in the sense that the out-of-gamut colors do not turn up in the colors measured from the target. By forming the gamut model, this implicit gamut information is made explicit.

In some of the embodiments, the printer is arranged and programmed to automatically carry out the following chain of activities, without manual user intervention: first, a target is printed. Then, the colors of the color-patch set are measured, either immediately or after an ink-drying delay (e.g. one hour), and the gamut model for that printed target is formed. In the case of an ink-drying delay, the medium with the target printed on it can be left in the printer (which is then at rest), and when the drying time is over, the printer will automatically resume operation by measuring the colors of the dried target. Some embodiments have the option that the target may be removed after the print and later be re-inserted for the color measurement and the construction of the gamut model, in order to enable the printer to be used, e.g. for other print jobs, in the meantime.

In further embodiments, the above automatic functionality is extended such that the printer can repeatedly carry out the chain of activities mentioned above for different print-medium/colorant/print-mode combinations. In these embodiments, a user may pre-select a sequence of different print media and/or print modes, and then start the procedure. The printer will then sequentially, for all pre-selected combinations print a target, measure its colors, and form the gamut model for this combination, without user invention. In some embodiments, the user has to put the different papers according to his pre-selection into the printer's paper tray before starting the procedure. In other embodiments, the printer is equipped with an automatic print-medium-type recognition so that the user can perform the print- medium pre-selection by simply putting a pile of the different print media of interest in the paper tray. In principle, the type of colorant used might also be varied automatically, if the printer is equipped with several automatically switchable ink reservoirs. However, a printer will usually not be equipped with switchable inks so that colorant variation will typically need the colorant to be exchanged manually. The automatic production of gamut models for different print-medium/colorant/print-mode combinations enables, in some embodiments, a comparative analysis of the gamut properties of these combinations to be made.

As mentioned above, some embodiments have the option that an already- printed target can be inserted in the printer to measure its colors and form the gamut model. The ability to process already-printed targets also enables a target printed by another printer to be color-measured by the embedded color measurement device, and a gamut model of the other printer's color gamut to be formed by the embedded color gamut constructor. Thus, for example, some embodiments are arranged to make a comparative analysis of the gamut of the present printer and the other printer.

In order to compare color gamut models, in some embodiments, the printer includes a gamut comparator. The gamut comparator may either be comprised in the printer, or in a printer-driving program module arranged to run in a computer connected to the printer, or in both. For example, the gamut comparator is arranged to compare two or more color gamut models associated with different print- medium/colorant/print-mode combinations of the present printer. Since, as mentioned, a target may also have been produced by another printer, the gamut comparator is also capable of comparing the same or different print- medium/colorant/print-mode combinations of the present printer and of the other printer. In some of the embodiments in which the gamut models are parameterized, the gamut comparison is made algorithmically by the comparator on the gamut parameterizations. In other embodiments, the comparator causes a display of the two (or more) different gamut models in an overlay manner, so that a user can examine the overlaid gamut models and compare them himself.

It may be necessary, for example, to compare two color gamut models of one and the same printer in order to characterize different print medium/colorant/print mode combinations with respect to the associated ability to reproduce as wide a range of colors as possible, and to select a certain combination. Comparing the color gamut of the present printer (for a certain print medium/colorant/print mode combination) with that of another printer may be motivated by what is referred to as "proofing". The printing of an image with a large printing press is normally only profitable for large quantities of printouts since print plates, or at least complex adjustments to the press, have to be made beforehand. Thus, the printing press will not be used for test prints at the image-design phase, but rather a "desktop" printer simulating the printing press will be used, also referred to as "proofing printer". To find out whether (or with what print medium/colorant/print mode combination) the present printer is able to simulate a certain printing press (again with a certain print medium/colorant/print mode combination), it is useful to compare the gamut model of the present printer with that of the printing press.

To this end, in some of the embodiments the comparator determines the intersection of the gamut model of the present printer and that of the other printer, in order to provide an indication whether the gamut of the other printer lies completely within the gamut of the present printer, i.e. an indication whether the latter is able to be used as a proofing printer for the other printer.

In some embodiments, the gamut intersection can be displayed graphically. In other embodiments, a percentage value is (also) provided which indicates the volume of intersection of the gamuts.

Up to this point, the term "gamut" has only been used in the context of a "device gamut", i.e. to characterize the range of colors that can be reproduced with the device considered (with a certain print medium/colorant/print mode combination considered). But the term "gamut" is also meaningful in connection with an image.

Such an "image gamut" characterizes the range of colors that occur in a certain image. Depending on the image considered, an image gamut may be small compared with a device gamut: for example, an image only showing human skin will only have skin tones occupying only a small volume in the color space. Whereas the formation of a device gamut model is based on a measurement of all the colors of a target with color patches spanning all the colors printable with the printer considered, the formation of an image gamut model may analogously be based on measurement of all the colors present in the image considered. Thus, by inserting a printed image (rather than a target) in the present printer and measuring its colors, the gamut constructor of some the embodiments is able to form a gamut model of that image. The printed image may, for example, be an image printed by a third printer. Alternatively, the image may be an image represented by digital image data, and the image gamut model may be formed using the digital image data. In embodiments with a gamut comparator, the gamut model of the present printer (with a certain print medium/colorant/print mode combination) can be compared with the image gamut model, and, for example, the intersection of the two models can be determined. This is useful, for example, for figuring out whether the image can be reproduced with the present printer (with the print medium/colorant/print mode combination considered).

For proofing applications, some of the embodiments are arranged to compare three gamut models: (i) the gamut model of an image (e.g. an image to be finally printed in a mass production); (ii) the gamut model of the present printer (with a certain print medium/colorant/print mode combination); and (iii) the gamut model of the printing press (again with a certain print medium/colorant/print mode combination). The comparator produces the intersection of the gamut models (ii) and (iii) and provides an indication as to whether the image gamut model lies completely within the intersection. A positive result indicates that the printing press is able to print the image, and the present printer is able to be used as a proofing printer for this (note that a positive result does not require the proofing printer to be able to emulate the printing press for all the colors that the printing press can print, but that it is sufficient that it can emulate the printing press for the image considered).

In some of the embodiments, the color stability of a target for a colorant/print medium combination is verified before the color gamut model is formed. This is motivated by the fact that, after a color patch has been printed on a medium, the color is generally not yet fixed but may still vary (develop) due to intrinsic ink properties and physical and chemical interaction of the ink with the print medium and the atmosphere. Thus, not only the ink and the print medium, but also external parameters, such as air humidity and temperature exert an influence on the color. Hence, it is useful to have a color stability parameter that indicates the time duration after which a color may be assumed to be stable (i.e. does not change anymore, or only shows changes below a certain threshold). Depending on the nature of the color-medium interaction, color stability may also depend on the print medium used. Therefore, in some of the embodiments device-independent color values of a set of color patches printed with the color printer are repeatedly measured, with the embedded color measurement device at different points of time. Color changes between the different points of time are calculated, and it is verified that color stability has been reached. The color gamut model for the colorant/print medium combination is then formed on the basis of a color measurement of a set of color patches which has been verified to be color-stable. The set of color patches may be a target which is finally used to form the gamut model. Alternatively, it may be a reduced set of color patches since less color gradation may be used for the stability measurement than for the gamut model construction. In some embodiments, the verification of color stability is based on a color stability parameter. The color stability parameter is based on color changes observed between the different points of time. To obtain a certain robustness against fluctuations, a determination of the color stability parameter is based on an averaging procedure, in some embodiments. For example, if one uses a hundred color patches (of the same color or different colors) as a basis for the stability verification, one assumes that color stability has been reached when the average change, averaged over the full 100-color sample, between measurements of the same color patch at different points of time (e.g. at two subsequent days) is below a threshold. Alternatively, if such an averaging procedure renders the color stability measurement too insensitive, one can require the maximum color change observed in the sample at the different points of time to be below a threshold. By means of the color-stability measurement, the system automatically knows when stability is reached for a certain print-medium/colorant combination, and can then construct the associated gamut model. Returning now to Fig.1 which shows a printing system 1 with a printer 2 and an external computer 14 to which it is connected. The computer 14 is equipped with a computer screen (graphical user interface) 19. The exemplary printer 2, for example a color inkjet printer, has four successively arranged print stations 3.1-3.4. A pile of print media 4 (e.g. sheets of papers) is stored in a paper tray 5 of the printer 2. A print medium 4 is automatically loaded into the print path by means of a paper feed mechanism and moved forward (the forward direction is indicated by an arrow) by means of a conveyor belt 6. Each of the four print stations 3.1-3.4 has an array of ink nozzles 7 arranged in widthwise direction of the print medium 4. Each print station 3.1-3.4 has a cartridge 8 filled with ink of one of the primary colors cyan (C), magenta (M), yellow (Y), and black (K) and is arranged to spray droplets of the corresponding ink onto the print medium 4 by means of the nozzles 7. The conveyor belt 6, which is arranged beneath the print stations 3.1-3.4, is guided by two rollers 10, wherein one of the rollers 10 is driven by an advance mechanism during the printing process to move the print medium 4 past the print stations 3.1-3.4. Owing to the spaced arrangement of the print stations 3.1-3.4, when the print medium 4 is conveyed, a certain point considered on the medium 4 passes the individual print stations 3.1-3.4 successively. In order to produce a multicolor image (in the example shown, a set of predetermined color patches, or target 9) with the four single-color images aligned, the print stations 3.1-3.4 correspondingly produce their single-color images in a timely shifted manner.

Cyan, magenta, yellow and black ink dots belonging to the same pixel are printed next to each other at different positions on the print medium 4 to ensure that the different inks do not overlap or intermix too strongly. A usual halftoning technique is used in the printer 2, for example a technique in which each dot is of the same size, and different densities, and thus colors, are obtained by printing a greater or smaller number of dots within a given area.

The target 9 is such that the predetermined set of color patches 9 covers all combinations of the four inks in a certain grid, e.g. 0%, 10%, 20%, 40%, 70%, 100% (thus, in this example, the target 9 has about 1 ,300 (6⁴) color patches).

After the target 9 has been printed, the print medium 4 is at rest for some time until the color becomes stable and is then further conveyed past a spectrophotometer 11 that is arranged to measure the LAB-values of the individual color patches of the target 9. A printer controller 12 with firmware has a gamut constructor firmware component 13.1 (being a part of a color gamut constructor 13). It is arranged to construct a part of a color gamut model, e.g. a parameterization of the gamut, based on the measured LAB-values of the target 9. The data representing the measured values of the target and the gamut parameterization are transferred to the computer 14 over an interface 15. The computer 14 has a printer-driving program module 16 which includes a color gamut constructor driver 13.2 being part of the overall color gamut constructor 13. The gamut constructor driver 13.2 is arranged to perform final operations to construct the gamut model, e.g. transform the gamut parameterization into a displayable representation which is then used to display the gamut model 33 on the computer screen 19. In alternative embodiments, the entire gamut constructor 13 is embedded in the printer 2, and in still further embodiments the gamut constructor 13 is entirely in the printer-driving program module 16 in the computer 14.

In some embodiments, the printer controller 12 is not only arranged to automatically carry out the above process of printing the target 9, measuring the color by means of the spectrophotometer 11 and constructing, in an ad-hoc manner, a gamut model by means of the gamut constructor 13 (or a part of it by means of the gamut constructor firmware component 13.1), but also to automatically carry this out for different print media/print mode combinations in a sequence, without manual user intervention. For example, different print media are automatically detectable in this sequence by a print-media-type sensor in the paper feed mechanism; so the user simply puts a pile of different print media 4 in the paper tray and starts the process, whereupon the sequence is carried out automatically.

In some embodiments, the printer controller 12 is also arranged to enable targets printed by the other printers (or by the present printer at an earlier stage) to be used. Such an already printed target can be inserted by a user, e.g. in the paper tray 5, and the paper feed mechanism then transports it, without print activity, past the print stations 3 to the spectrophotometer 11 that measures the target's color. The gamut constructor 13 is then also able to construct a gamut model for such an "external" target. In some embodiments, the gamut constructor 13 is also arranged to compare gamut models of different targets (for example of a target 9 printed with the present printer 2 and an external target), and to prepare a representation of the results of the comparison, e.g. a graphical representation of those parts of the gamut volume of the external target that are out of the present printer's gamut. Furthermore, the printer controller 12 has a color-stability firmware component

17 arranged to calculate differences between the colors of the color patches of the target 9 measured at different points of time and to determine a color stability parameter, as will be explained in more detail below.

Since the color reproducible by the printer 2 may not only depend on the ink type, the print medium or the print-mode used, but also on environmental conditions, such as temperature and humidity, the printer 2 is also equipped with suitable sensors 18, such as a thermometer and a hygrometer. Their measurement values are transmitted to the printer controller 12 and the computer 14.

Fig. 2 is a high-level diagram of how a color gamut model construction is invoked from an application program and processed. Two exemplary applications 20,

21 are shown here, which need, as an input, a gamut model of the printer 2 (and, optionally, of other external printers).

The applications 20, 21 reside on the application layer of the computer 14, and not on its operating system layer (the distinction between the "application layer" and the "operating system layer" typically depends on the specific operating system used - for example, in UNIX-type operating systems the application layer refers to what is called the "user mode", and the operating system layer to the "kernel mode").

The first exemplary application 20 is a color-catalogue customizer. "Catalogues" of colors, such as the Pantone color set, are collections of discrete colors which populate the color space with a certain density and spacing, and are arranged in a certain structured manner in the catalogue so that they can easily be worked with by graphics designers. If a certain printer has a reduced gamut and is thus not able to reproduce a whole color catalogue, there is sometimes a need to map the color catalogue into the printer's gamut in a manner, however, in which the structure of the catalogue, e.g. the relative arrangement of the discrete colors in color space, is maintained. This mapping is also called "customizing" the color catalogue. An application that performs the mapping is called, herein, a "color-catalogue customizer". Since the actual mapping to be performed normally depends on the gamut of the printer, the color-catalogue customizer 20 typically has a need to get the color-gamut model of the printer (here the printer 2) to perform its task to map the color catalogue.

The second exemplary application program 21 is a CAD program. Although gamut models can also be displayed by a graphical interface on the operating system level, there may be a need to display, use and modify a gamut of a printer in a CAD program. To this end, the CAD program 21 also needs to get the color-gamut model of the printer 2 to perform its tasks.

As usual, the computer 14 has an operating system 22 on the operating system layer. On the operating system layer, and linked with the operating system 22, is the printer-driving program module 16. It includes a usual printer driver 23 and the color-gamut constructor driver 13.2. Below the operating system 22 and the printer-driving program module 16 is a communications port driver 24.

The application 20 or 21 generates a request to obtain a gamut model of the printer 2 and sends this, like a usual request for an operating-system service, to the color-gamut constructor driver 13.2. The constructor driver 13.2 passes it via the communications port driver 24 to the printer 2. This causes the printer 2 to print the target 9, measure the target's colors with the spectrophotometer 11 , construct, in an ad-hoc manner, a first part of the printer's gamut model (e.g. a simple parameterization of it), and send this back to the communications port driver 24 which passes it to the color-gamut constructor driver 13.2. The constructor driver 13.2 finishes the construction of the printer's gamut model (e.g. transforms the parameterization into a displayable parameterization) and returns it to the invoking application 20 or 21. In this architecture, application programs, such as 20 and 21 , are enabled to invoke an ad-hoc gamut model construction of the printer(s) connected to the computer 14, in the usual form in which an operating system service is invoked, such as a request to print out a document.

This is illustrated in Fig. 3 by an exemplary user interface that enables a user of the application 20 or 21 to invoke printer services from within the application 20 or 21. The user interface shown is a printing menu 25 which enables a user to select a printer from a set of printers connected to the computer 14, here, two HP inkjet printers and two HP laserjet printers, by clicking at a pointer in a printer selection bar 26. An input field 27 is also provided which enables the user to specify the pages she/he wishes to print. A button "Options" 28 enables a user to further specify options of how the document is printed out. Check boxes 29 enable a print model to be chosen. Finally, by clicking a button "color gamut construction" 30 the user can trigger an ad-hoc color gamut construction, as described in connection with Fig. 2. Fig. 4 illustrates the color-catalogue customizer 20 of Fig. 2. It maps an original color catalogue 31 , also called a "swatch book", into colors reproducible by the color inkjet printer 1 on a certain print medium, etc., thereby forming a swatch book 32 customized to the printer 2 and the print medium, etc. The colors of both swatch books 31 , 32 are represented as digital color values, e.g. in the LAB color space. The original swatch book 31 has typically colors that cannot be reproduced by the printer 2, i.e. that are out-of-gamut. The color-catalogue customizer maps the colors of the original swatch book 31 to colors that are all in gamut. To this end, the color-catalogue customizer typically needs a color-gamut model of the printer 2 with the print medium in question, etc. To produce the customized swatch book 32, the color customizer 20 determines, in an ad-hoc manner, a color gamut model of the color inkjet printer 2 with the specific print medium, by submitting a corresponding request to the color-gamut constructor 13 (Fig. 2). As explained above, a target 9 is printed, measured, and a gamut model 33 is constructed in an ad-hoc manner and returned to the color-catalogue customizer 20. On the basis of the gamut model 33, the customizer 20 knows how to map the colors of the original swatch book 31 , so that they all become in-gamut-colors.

Fig. 5a illustrates that the gamuts of two different printers, or of the same printer with different print media, will generally be different, and that these differences show up in the gamut models constructed. It shows two exemplary color gamut models produced by the color gamut constructor 13 and represented by a gamut viewer which is, for instance, also provided by the color gamut constructor driver 13.2 (Fig. 2) and displayed on the screen 19 (Fig. 1 ). For example, the same CMYK color values have been sent to the color inkjet printer 1 to print the target 9 twice, on the one hand on coated paper, on the other hand on glossy paper. The targets 9 are measured by the spectrophotometer 11 and the corresponding LAB-color values are stored as data points. The construction of the color gamut model, as explained, e.g., in US2002/0140701 A1 , involves, for example, the construction of a convex hull over the color data points in an additive color space. A suitable additive color space is, for example, the CIE-XYZ color space. A reason for constructing the convex hull in an additive color space is that the vertices can be connected with straight lines, since one can expect all the colors in an additive color space to be generated by different combinations of two given colors to fall on a straight line connecting the two given colors. To construct the hull, the measured LAB-values are transformed into XYZ. Once this convex hull is constructed in the additive color space, it is, for example, transformed back into a corresponding object in a psychometric color space, and presented as the color gamut model to the user or an application program. Suitable psychometric color spaces include, for example, the CIE-LAB color space and the CIE-LUV color space. In Fig. 5a, color gamut model 33' for the coated paper is represented as a solid model whereas that for the glossy paper is requested as a wireframe model, so that both models can be overlaid, and the difference between the two overlaid gamut models 33', 33" is still visible. A similar picture may represent the different gamuts of different printers, or that of different colorants.

Fig. 5b shows another example of two different gamut models produced by the color gamut constructor 13 and displayed on the screen 19. Here, the smaller gamut model represents a color gamut of an image, i.e. the set of all colors appearing in the image. The larger gamut is the printer's gamut. The image gamut can be constructed by measuring the (e.g. externally) produced image with the spectrophotometer 11. If the image gamut model lies entirely within the printer's gamut model as shown in Fig. 5b, the particular image can be reproduced by the printer 2, even if the image was printed on a press that has a larger gamut than the printer 2 and, thus, cannot be entirely simulated by the printer 2.

Since the ad-hoc ability of the printing system 1 to construct gamut models is particularly useful for a "proofing printer", Fig. 6 illustrates proofing. As mentioned at the outset, no two types of color capturing or reproducing devices will operate in exactly the same color space, which means e.g. that if two digital cameras (produced by different manufacturers) take a photo of the same object (under the same lighting conditions), then the sets of RGB color values produced which represent the captured image will be different. Therefore, a manufacturer of digital cameras will normally provide an input profile by means of which measured RGB-color values are transformed into a device-independent color space, such as device-independent RGB or LAB. The same is true for output devices: For example, a certain set CMYK values, when reproduced by printers of different types (or on different print media etc.), will give different colors. Therefore, each printer, or printer driver, is normally equipped with output profiles that correct this. Furthermore, the output profiles normally include a color mapping to transform out-of-gamut colors into reproducible colors, e.g. according to one of the known rendering intents mentioned at the outset.

RsGsBs- and RcGcBc-color spaces 50, 51 of a scanner 71 and a camera 73 are shown, and input profiles 60, 61 (implementing an AtoB-mapping) represent a transformation from the device-dependent RsGsBs- and RcGcB_c-color spaces into the device-independent LAB-color space 58 (since the different color devices, although operating with RGB- or CMYK-values, actually refer to different color spaces, the color spaces are distinguished by indices).

On the output side, a monitor 75 has an R_MG_MB_M-COIOΓ space 52, and by means of an output profile 62 (BtoA-mapping), colors are mapped from the LAB-color space 58 to the R_MG_MB_M-COIOΓ space. As mentioned above, the mapping usually includes a transformation of out-of-gamut colors into reproducible colors according to a rendering intent.

The printer 2 (Fig. 1) also has its own color space 53, which is, for example, a CMYK-color space (for the sake of simplicity, the fourth axis "K" is not depicted in

Fig. 6). An output profile 65 (BtoA-mapping) maps colors from the LAB-color space 58 to the C|M|Y|Kι-color space (the index "I" stands for "inkjet"). Analogously, a printing press 77 has a CpM_PY_PKp-color space 54. An output profile 67 (implementing a BtoA-mapping) maps colors from the LAB-color space 58 to the CpM_PY_PKp-color space. As mentioned above, the output mappings usually include a transformation of out-of-gamut colors into reproducible colors according to a rendering intent (often several different mappings are provided, for the different rendering intents).

Furthermore, in Fig. 6, an "input" profile 66 (AtoB-mapping) of the printing press 77 is also shown, although the printing press 77 is only an output device. This input profile 66 is useful for "proofing" the printing press 77 by means of another device. "Proofing" means simulating the reproduction of images of an output device which is not readily available for test prints (such as the printing press 77) by another output device, e.g. by a monitor (this would be called "soft-proofing") or another printer, such the printer 2, which is, e.g., a desktop inkjet printer ("hard-proofing").

To simulate the printing press 77 on the proofing printer 2, a three-stage mapping procedure is carried out: first, the image to be simulated is mapped from the device independent LAB-color space 58 into the printing press's C_pMpY_PKp - color space 54 by means of the output table 67, as if it were to be printed by the printing press 77. As mentioned above, this includes a mapping of out-of-gamut color (referring to the printing press's gamut) into in-gamut color, i.e. colors that can be reproduced by the printing press. Then, instead of actually printing with the printing press 77, the image - which is now represented in the C_pMpY_PKp - color space 54 is transformed back to the LAB-color space 58 by means of the input profile 66. However, since all the colors can be represented in the device-independent LAB- color space 58 (it has no limiting gamut) no such "gamut-mapping" is now performed in the backward mapping, but the colors are only transformed into the LAB-color space, without any color change. The result of this is a LAB-representation of the image as it would be printed by the printing press 77. At the last stage, this LAB- representation is transformed into the printers C|M|Y|K| - color space 53 by means of the profile 65. If the printer's gamut is smaller than the printing press's gamut, this last mapping will include another "gamut mapping" of the out-of- gamut colors (now referring to the printer's gamut); i.e. the printer 2 will not be able to (perfectly) simulate the printing press 77. The ad-hoc gamut determination functionality described is useful in such a proofing context: if the printer's gamut does not circumscribe the printing press's gamut, the printer cannot perfectly simulate the printing press (with the particular print medium/colorant/print mode combination considered). By determining the printing press's gamut model (e.g. by using a target at the printing press) and the printer's gamut models for different print media/colorant/print mode combinations and comparing the latter with the printing press's gamut based on the gamut-comparison functionality described, the user can, for example, find out what print medium is to be used in order to be able to simulate the printing press with the printer, by requiring that the printer's gamut with this print medium circumscribes (or nearly circumscribes) the printing press's gamut.

This is illustrated in Fig. 7 which again shows an exemplary output of a gamut viewer (that is part of the gamut constructor 13) on the screen 19. To keep Fig. 7 simple, it shows different gamut models as two-dimensional objects (normally, these models are three-dimensional in LAB units). Three gamut models are shown: the gamut model of the printing press 77 to be simulated (associated with a certain paper, e.g. coated paper), and two gamut models of the proofing printer 2, associated with two different paper types, e.g. glossy paper and coated paper. Only one of the proofing printer's gamut models (the one associated with glossy paper) circumscribes the printing press's gamut model (so that the latter lies completely in it), whereas the other proofing printer's gamut model (the one associated with coated paper) does not completely circumscribe the press's gamut model. In other words, the proofing printer can simulate the printing press when glossy paper is used by the proofing printer for the simulation points, but it cannot entirely simulate the press when it uses coated paper for the simulation.

The ability to produce gamut models of different print media/colorant/print mode combinations including gamut models of other devices, such as printing presses, and to compare them with one another in an ad-hoc manner enables the most suitable of these combinations for the simulation to be chosen by the user, in a simple, fast and reliable manner. In addition to such a graphical gamut model comparison, the printing system is arranged to compare the gamut models in a quantitative manner. For example, the quotient of (i) the volume of the intersection of the gamut model of the proofing printer with the gamut of the printing press and (ii) the color gamut volume of the printing press is calculated by the gamut comparator and displayed on the screen at 34. This quotient is referred to as "press coverage" In the example shown, the press coverage of the proofing printer 2 with glossy paper is 100% (values of the quotient greater than 100% are set to 100%), whereas the press coverage of the proofing printer 2 with coated paper is 94%. The larger the press coverage is, the better the printing press 77 can be emulated by the proofing printer 2.

As illustrated in connection with Fig. 5b, if only a certain image is to be proofed by the proofing printer 2, a comparison between the image gamut model (as it is to be printed on the printing press) and the proofing printer's gamut models for different media/colorant/print mode combinations, can also be made by the gamut comparator. If, for example, an image with only skin tones is to be printed, and if the gamut model of this image lies entirely within the proofing printer's smaller gamut model with coated paper, such coated paper will allow a press-and-image coverage of 100%, and could also be used for proofing the image. Fig. 8 is a flowchart further illustrating the ad-hoc gamut model comparison of

Fig. 7 for one particular medium/colorant/print mode combination. At 100, a target is printed by the printer 2 on the print medium (e.g. glossy paper) using the colorant and print mode. At 101 , the LAB-values of the color patches in the target 9 are measured with the embedded spectrophotometer 11. A color gamut model is determined for the printer 2 and medium/colorant/print mode combination used when printing the target 9 at 102. At 103, the LAB-values of color patches on a target printed by a printing press 77 are determined. To this end, the externally printed target (printed by the printing press 77) is inserted into the printer 2, and the LAB- values of these color patches are measured with the embedded spectrophotometer 11. Based on these LAB-values, a color gamut model for the printing press 77 is determined at 104. At 105, both color gamut models are displayed by the gamut viewer on the screen 19. At 106, the press coverage is calculated and also indicated on the screen 19. Of course, the sequence 103-104 can be carried out before 100- 102. Furthermore, in some embodiments, the LAB values of the printing press's target are not measured at 103, but an already existing color mapping profile for the printing press is used instead of the printing press's gamut model as determined at 104.

Fig. 9 is a similar flowchart illustrating automatic ad-hoc gamut model construction for different media, e.g. different types of paper. At 110, sheets of paper of different types are inserted into the printer 2. At 111 , the upper sheet is fed to the conveyor 6, the paper type is automatically recognized by a paper type sensor, and the sheet is automatically drawn in and the target 9 is printed on it. The LAB-values of the target's color patches are measured with the embedded spectrophotometer 11 at 112. At 113, a color gamut model is determined for the measured LAB-values. At 114, it is ascertained whether the paper tray is empty. If the answer is negative, the activities 110 to 113 are repeated. If the answer is positive, the different gamut models are displayed, compared with one another and/or with other gamut models (e.g. of a printing press) etc. at 115, to enable the user to select the paper which is best-suited for the purpose intended, e.g. to simulate a certain press and/or a certain image on the printer.

Some embodiments of the printing system 1 (Fig .1 ) have an additional functionality performed by the color-stability firmware component 17 pertaining to color stability. The colors of a printed image are generally not stable over time, and for certain print media/colorant combinations this is a significant effect. In particular, the colors may significantly change in a time period after printing. In order to provide precise gamut analyses (such as the gamut comparisons of Figs. 5 and 7) in such situations, some embodiments of the printing system 1 are also arranged to determine color stability. For example, this allows the printing system, after having printed a target 9, to assess whether color stability has already been reached, and to perform a gamut model construction and analysis, as described above, on the stable colors.

The term "color stability" refers to the color change of an image from the point of time of the printout until its color changes are no longer visually perceived. To control the measurement, only the color changes due to causes other than exposure to light are considered, i.e. only color changes occurring to printed samples while they are stored in the absence of light are considered. However, it is also possible to expose the image to light and to measure the color changes.

Fig. 10 is an exemplary color-stability diagram. The vertical axis represents the color change (in terms of ΔE) of a target's printed color patches. By definition, 1 ΔE unit is the change in color which can just be perceived by the human eye. All color changes smaller than 1 ΔE cannot be perceived by the human eye. The color changes between two color patches of the same color measured at different points of time may be obtained by means of the following formula which calculates the Euclidean distance between two colors in the LAB color space.

AE_a ^* _b = V '(ΔL^*)² + (Δa^*)² + (AbJ The horizontal axis represents the time (in hours) passed since the target 9 was printed. A second curve indicates the ΔE change over time with regard to the maximum color change of the color patches. It should be mentioned that the measurements are done always comparing to the zero point stage, i.e. with regard to the point of time of the printout, so that the value for ΔE at the different points of time is a cumulative indication.

It can be seen that, in the example shown in Fig. 10, after three hours, the maximum color changes fall below the perceivable visible threshold of one 1 ΔE. Thus, it is reasonable to assume that the colors in the target are stable after three hours and the printout may then be used for the gamut model determination described above. The average color changes are smaller. The diagram also indicates that the main color change occurs during the first minutes or hours after printing. The color stability performance depends on the medium/colorant combination. A determination of the color stability firmware component 17 therefore not only ensures that the gamut model is only determined when color stability has been reached, but also provides the user with a further criterion for selecting a certain medium colorant combination (since the user might be interested in using a combination which achieves the color stability rapidly).

Fig. 11 is a flowchart of the activities governed by the color stability firmware component 17 (Fig. 1 ) to ensure that an ad-hoc gamut model determination is only performed after color stability has been reached. At 120, a target 9 is printed by the printer 2. At 121 , the LAB-values of the color patches, or some of the color patches, in the target 9 are measured with the embedded spectrophotometer 11 , at predefined intervals, e.g. every 15 minutes (in other embodiments, the intervals between the measurements are shorter in the beginning and are increasingly spread apart). The printed target may remain in the printer 2, under the spectrophotometer 11 , e.g. when the stability is determined overnight, but it may also be temporarily removed from the printer 2 between the measurements, if the printer 2 is to be used for other print or gamut-determination jobs. At 122, the LAB-value of each measured color patch is compared with the corresponding LAB-value of the previous measurement. From this comparison ΔE (it may, e.g., be the maximum of all the color patch's ΔEs, or the average of all the ΔEs, as explained in connection with Fig. 10) is determined. Since there may be some fluctuations if the ΔE is only determined from a comparison with the previous measurement, in some embodiments a combined ΔE is calculated from multiple previous measurements (e.g. the average of the (maximum or average) ΔEs from the three previous measurements). At 123, it is ascertained whether the ΔE calculated (in one or the other way) is below a predetermined threshold, e.g. below one (in some embodiments, it may be required that it be repeatedly below the threshold, e.g. three times). If the answer is negative, further color measurements and comparisons are made. If the answer is positive, it is assumed that color stability has been reached, and the color gamut model is determined at 124, in the manner described in connection with Figures 1 to 11 above, for example using the last color measurement (if all the color patches were measured), or on the basis of a further color measurement to be performed. Although the entire process may take some hours, or even days, it is considered as an ad-hoc gamut determination, since the delays are caused by the time intrinsically needed by the colors to become stable, rather than by handling delays, or the like, and the whole process is carried out automatically (it is carried out automatically without any user interaction if the target remains in the printer; if not, the user, for example, may restart the automatic process when he/she has re-inserted the target into the printer). The functionality to determine when color stability has been reached can also be used without subsequent gamut-construction, if the main interest is not the determination of a (color-stable) gamut model, but the determination per se of when color stability is reached. As a result of such a determination, a stability diagram like the one shown in

Fig. 10 is produced and displayed.

Thus, the embodiments described enable a color gamut model of a printer for a certain print medium/colorants/print mode combination to be determined in an ad- hoc manner. All publications and existing systems mentioned in this specification are herein incorporated by reference.

Although certain methods and products constructed in accordance with the teachings of the invention have been described herein, the scope of coverage of this patent is not limited thereto. On the contrary, this patent covers all embodiments of the teachings of the invention fairly falling within the scope of the appended claims either literally or under the doctrine of equivalents.

Claims

What is claimed is:

1. A printing system, comprising: a printer arranged to render color images on print media and having a color gamut; a color measurement device embedded in the printer and arranged to measure colors, printed by the printer on a print medium, in a device-independent color space, and a color gamut constructor arranged to form, in an ad-hoc manner, a model representing the printer's color gamut with its boundary, based on a color measurement by the color measurement device of a set of color patches printed by the printer, wherein the color gamut constructor is comprised in the printer or in a printer- driving program module arranged to run in a computer connected to the printer, or in both.

2. The printing system of claim 1 , wherein the printer's color gamut depends on at least one of: the print medium used, a colorant used, and a print mode used, so that the gamut model formed will be specific for the print-medium/colorant/print-mode combination used.

3. The printing system of claim 1 , arranged to automatically carry out the following chain of activities, without manual user intervention: print a set of color patches, measure the colors of the color-patch set, and form the gamut model for that printed color-patch set.

4. The printing system of claim 3, arranged to repeatedly carry out the chain of activities for different print-medium/colorant/print-mode combinations.

5. The printing system of claim 1 , arranged to also enable a color-patch set printed by another printer to be color-measured by the embedded color measurement device and a gamut model of the other printer's color gamut to be formed by the embedded color gamut constructor.

6. The printing system of claim 1 , further comprising a gamut comparator arranged to compare, or to enable a user to compare, two or more gamut models associated with different print-medium/colorant/print-mode combinations of the printer, or with the same or different print-medium/colorant/print-mode combinations of different printers.

7. The printing system of claim 1 , wherein the gamut constructor is also arranged to form a gamut model of an image to be printed, and wherein the printer further comprises a gamut comparator arranged to compare, or to enable a user to compare, the image's gamut model and a device gamut model.

8. The printing system of claim 7, wherein the device gamut model is the printer's gamut model measured and formed for a certain print-medium/colorant/print- mode combination, or a gamut model of another printer, or a combination of both.

9. The printing system of claim 1 , further arranged to enable the gamut constructor to be invoked from a gamut-model-processing computer application to form the gamut model, and the gamut constructor to return the gamut model formed to the gamut-model- processing computer application.

10. The printing system of claim 1 , further arranged to verify color stability of a colorant/print medium combination for which a color gamut model is to be formed, by repeatedly measuring device-independent color values of printed colors with the embedded color measurement device at different points of time, calculating color changes between the different points of time, and verifying that color stability has been reached; forming the color gamut model for the colorant/print medium combination based on a color measurement of a set of color patches which has been verified to be color-stable.

11. A method of forming a device gamut model, comprising: printing a set of color patches with a printer on a print medium; measuring the colors of the set of color patches in a device-independent color space by a color- measurement device embedded in the printer, and forming, in an ad-hoc manner by a color gamut constructor, a model representing the printing device's color gamut with its boundary, based on the color measurement of the set of color patches, wherein the color gamut constructor is comprised in the printer or in a printer- driving program module arranged to run in a computer connected to the printer, or in both.

12. The method of claim 11 , wherein the printer's color gamut depends on at least one of: the print medium used, a colorant used, and a print mode used, so that the gamut model formed will be specific for the print-medium/colorant/print-mode combination used.

13. The method of claim 11 , wherein the activities of printing, measuring and forming are automatically carried out in a sequence, without manual user intervention.

14. The method of claim 13, wherein the activities of printing, measuring and forming are carried out repeatedly for different print-medium/colorant/print-mode combinations.

15. The method of claim 11 , comprising: measuring a color-patch set printed by another printer by the embedded color measurement device, and forming a gamut model of the other printer's color gamut by the embedded color gamut constructor.

16. The method of claim 11 , comprising: comparing, or enabling a user to compare, two or more gamut models associated with different print-medium/colorant/print-mode combinations of the printer, or with the same or different print-medium/colorant/print-mode combinations of different printers.

17. The method of claim 12, comprising: forming a gamut model of an image to be printed by the embedded color gamut constructor, and comparing, or enabling a user to compare, the image's gamut model and a device gamut model.

18. The method of claim 17, wherein the device gamut model is the printer's gamut model measured and formed for a certain print-medium/colorant/print-mode combination, or a gamut model of another printer, or a combination of both.

19. The method of claim 11 , further comprising: invoking the gamut constructor from a gamut-model-processing computer application to form the gamut model, and returning, by the gamut constructor, the gamut model formed to the gamut- model-processing computer application.

20. The method of claim 11 for verifying color stability of a colorant/print medium combination for which a color gamut model is to be formed, further comprising: repeatedly measuring, with the color measurement device at different points of time, device-independent color values of a set of color patches printed with the color printer, calculating color changes between the different points of time, and verifying that color stability has been reached; forming the color gamut model for the colorant/print medium combination based on a color measurement of a set of color patches which has been verified to be color-stable.