Appendix Introduction Contents of the Appendix 1. Visual timeline 2. Glossary of terms 3. An overview of halftoning and digital print technologies

Transcription

1 Appendix Introduction The chapters contained in the appendix are presented as background information to support the main body of the thesis. The appendix includes the technical detail of early experiments and background reference data for the reader. The appendix also includes a visual timeline of artists and technologists colour models, which are referenced in the body of the thesis. The timeline demonstrates the different methods of colour classification, from those using lines and circles to two and three dimensional models and diagrams, which can be read from left (prehistoric models) to the right (contemporary models). As a demonstration of dissemination and knowledge exchange, there is also a list of meetings, publications, papers and workshops that were undertaken during the period of research. Contents of the Appendix 1. Visual timeline 6 2. Glossary of terms 7 3. An overview of halftoning and digital print technologies Traditional Halftoning Stochastic Halftoning Basic types of digital print technologies How the basic digital print processes work Drop on demand Industrial drop on demand How a drop is formed in Piezo head technology A description of existing inkjet colour systems A description of parallel experimentation into colour accuracy, papers and inks and longevity 4.1 Accurate methods for printing colour 4.2 Generating the colour circle Inkjet printing the colour circles Preparing and making the colour separations for screenprint Results Conclusion Inkjet papers and inks Early pilot study with Encad Novajet printer

2 4.3.2 Results of the Novajet experiment Conclusions Ink on paper trials using the HP5500ps Ink on paper trials using the HP 5500ps ( ) initial 44 results Ink on paper trials, results after refinement Ink on paper trials, conclusions The longevity of papers and inks An assessment of appropriate methods for testing Measuring colour changes during exposure to light under 54 magnification Photographing the samples Results from measuring and photographing the samples Other indications that affect colour changes Conclusion to the longevity of papers and inks Conclusions to chapter Other experiments relating to Chapter 4 of the thesis: The 65 development and visualisation of bespoke colour sets for inkjet printing 5.1 Early experiments into the creation of colour charts Printer hardware and settings Colour charts and three-dimensional plots How and where to install the profiles Using the profiles 5.4 A description of the colour test charts Template 1: Blended squares Template 2: Blended six-pointed stars Template 3: To assess the appearance of printed colours 119 through the overprinting of solid colours Method for constructing template Method for printing the template Measuring the spectral reflectance density of colour Template 4: To assess the appearance of printed colours 121 through printing patterns on top of solid colours 5.5 Visualising and comparing the mixed inks: Results from chapter of the main thesis 6. Supporting documentation for Chapter 5: An artist s 128 introduction to inkjet: A case study and a generic working model 6.1 Version 1 of print 1 Hub Cap: FRD 857, November Version 2 of print 1 Hub Cap: FRD 857 December Print 2 version 1, Caprice CHV954, January Print 2 version 2, Caprice CHV954, January Supporting documentation and results of developing a custom 131 inkjet palette, measuring the third colour chart and comparing the results Bibliography 136 2

3 List of Figures Fig. 1 An example of simultaneous contrast 13 Fig. 2a Stochastic patterns 16 Fig. 2b Conventional and stochastic halftoning 16 Fig. 3 Inca Digital, The Columbia Turbo 17 Fig. 4a Continuous ink jet system 19 Fig. 4b Drop on demand ink jet system 19 Fig. 5 How a droplet is formed 21 Fig. 6 Two colour circles generated in Hue Saturation and Brightness. 26 Fig. 7 HSB slider in Photoshop. 27 Fig. 8 Primary hue inks screenprinted onto white card. 28 Fig. 9 The spectral data of the screenprinted colours. 29 Fig. 10 L*a*b* colour measurements, from the screenprinted ink 29 patches, are added to generate a custom CMYK profile. Fig. 11 The original HSB circles in LAB mode, in comparison to the 32 same circles converted to a generic CMYK colour mode. Fig. 12 These two circles are converted using different CMYK profiles. 32 Fig. 13 Colour circle rotated and flipped to the a* b* axis of the 33 measured charts. Fig. 14a Screenprint hues. 33 Fig. 14b HP Designjet 130nr. 33 Fig. 14c Screenprinted colours from colour separation. 33 Fig. 14d Z3100 using an output paper profile. 33 Fig. 14e Z3100 output same as source. 33 Fig. 15 Comparing two profiles that were used to convert to a CMYK 34 colour mode. Fig. 16 The first test page composed for the Encad Novajet. 39 Fig. 17 Test sheet for the HP 5000ps, comprising digital images and 43 text. Fig. 18 All the prints were arranged on a table and compared based 44 on a series of criteria Fig. 19 Particular areas that were compared during the refinement of 47 the selection process Fig 20a and 20b Printed colour inks onto a range of coated and 54 uncoated papers Fig. 21a Tosa Shoji 50 % black Fig. 21b Rayon Unryu 100 % black 58 Fig. 22a 50% cyan dye onto fine art coated 59 Fig. 22b 50% cyan pigment onto fine art coated 59 Fig. 22c 50% cyan dye onto fine art coated after 11months 59 Fig. 22d 50% cyan pigment onto fine art coated after 11 months 59 Fig. 23a 50% cyan dye onto fine art uncoated 59 Fig. 23b 50% cyan pigment onto fine art uncoated 59 Fig. 23c 50% cyan dye on onto fine art uncoated after 11months 60 Fig. 23d 50% cyan pigment onto fine art uncoated after 11months 60 Fig. 24a 50% magenta dye onto fine art coated 60 Fig. 24b 50% magenta pigment onto fine art coated 60 Fig. 24c 50% magenta dye on onto fine art coated after 11months 60 Fig. 24d 50% magenta pigment onto fine art coated after 11 months 60 Fig. 25a 50% magenta dye onto fine art art uncoated 61 3

4 Fig. 25b 50% magenta pigment onto fine uncoated Fig. 25c 50% magenta dye on onto fine art uncoated after 11 months Fig. 25d 50% magenta pigment onto fine art uncoated after 11 months Fig. 26a 100% Black dye on fine art coated Fig. 26b 100% Black dye on fine art coated after 11 months. Fig. 27 Some examples of the charts produced using the Dr Johnsons inks Fig. 28 HP130nr, 24 printer. 67 Fig. 29 An example of a printed colour chart using HP mixed inks 68 Fig. 30a and Fig. 30b By swapping the cyan and the black cartridges 68 different colour effects can be achieved Fig. 31a - 31c By moving the premixed colours to different channels 69 different colour mixtures can be obtained. Fig Fig. 36 Examples of a more systematic approach to ink mixing and refinement of the process Fig. 37 List of profiles in the ColorSync folder Fig. 38c The procedure for assigning a profile to soft preview the colours. Fig. 39 Three variations of CMYK blends. Fig. 40 The creation of the blended stars in Photoshop. Fig. 41 The layers option is used to print the separate layers of colour Fig. 42 An example of overprinting of coloured squares using pink, deep yellow and green primary hue colours. Fig. 43 A graded square with the four primary colours at each corner: 123 cyan, magenta, yellow and black Fig. 44 Example of the patterns that are printed over solid colours 123 Fig. 45 Composite layers in Photoshop of version 1 of 128 Hub Cap: FRD 857 Fig. 46a Version 1 Caprice CHV954 Fig. 46b Version 2 Caprice 130 CHV954 Fig. 47a Spectral comparison of the Deep Yellow, Yellow inks, I8 and 132 I18 Fig. 47b Spectral comparison of the Red ink, I1, I10 and I Fig. 47c Spectral comparison of the Blue ink, Y24, X24 and Y Fig. 47d Spectral comparison of the Green ink, H17, A8 and A9 134 Fig. 47e Spectral comparison of the Ultramarine ink, X22, X21 and X13 Fig. 47f Spectral comparison of the Turquoise ink, 2H12, 2D11 and 2D12 List of Tables Table 1. Technology comparison: Drop on Demand (DOD) 18 Table 2. Technology Comparison: Continuous Ink Jet (CIJ) 19 Table 3. Results of the first Novajet experiment Table 4. Ink on paper trials using the HP 5500ps ( ) Table 5. Ink on paper trials, results after refinement 48 Table 6. Average E differences after 11 months between the original 56 dye and pigment print samples and the exposed samples. Table 7. A description of the changes to the colour patches over time

5 Table 8. In order to maintain parity for all prints, the same working 67 space and printer settings were used Table 9. System folders on other operating systems 116 Table 10. A comparison of the L*a*b* values of the acrylic swatches 131 against the measured inkjet patches. List of Colour Charts Chart 1. rose-violet-dkyellow 72 Chart 2. rose-violet 76 Chart 3. red-violet 80 Chart 4. blue-green-black 84 Chart 5. red-green-black 88 Chart 6. red-green 90 Chart 7. rose-green 94 Chart 8. red-blue 98 Chart 9. rose-blue 102 Chart 10. green-blue 106 RUTH 110 5

6 1. Visual timeline 6

7 2. Glossary of terms The following section provides brief explanations of colour and technical references that might be unfamiliar to the artist. It is expected that many of these terms have been explained in the main body. For ease of reference these terms are listed in alphabetical order. a*: red-green coordinates in the CIEL*a*b* colour space. The positive value tends towards redness, whereas the minus values indicates colours moving towards green. Additive Colour Mixture: mixing of the three primary colour light (red, green and blue) to obtain colours. For example, combining green and red light creates yellow, red and blue creates magenta, and blue and green creates cyan. Appearance: a visual sensation through which an object is perceived to have attributes such as size, shape, colour, texture, gloss, transparency and opacity. (Mike North, 2009) 1 Assimilation: a perceived shift in a colour towards its neighbouring colour. The influence of neighbouring colours is mutual. The simplest example is the effect of a coloured pattern against a white background. The white will tend towards the neighbouring colour, and mutually the neighbouring colour will be lightened by the white. b*: yellow-blue coordinate in the CIEL*a*b* colour space. The positive value tends towards yellowness, whereas the minus values indicate colours tending towards blue. Blue Wool Scale: a method of comparing eight blue dyed strips of wool, that have different lightfastness properties, 1 being the most sensitive and 8 being the most resistant to fading. Blue wool has long been used by the 1 Definitions presented by Mike North to researchers at the 4 th CREATE training event, Hungary, Subsequent references are abbreviated to MN. 7

8 museum sector and textile industry, but more recently because of its known fading parameters, is also used by the print industry; it is commonly seen in historic houses, for example, the National Trust use blue wool cards to check the effect of sunlight on textiles. The blue wool cards are placed alongside the sample to be tested. Part of the blue wool is covered, which is used as a comparison at the end of the testing period. At the end of the test period, the blue wool samples are compared to the faded samples and given a light fastness rating that corresponds to the change in the blue wool; the higher the number the more lightfast the sample. Chroma: (see Saturation) CIE: Commission Internationale de l Eclairage the international Commission on Illumination CIE L*a*b* (or CIELAB or LAB): a perceptually uniform colour space in which the three L*a*b* values are plotted in a three dimensional space. It is a useful method for measuring reflective and transmissive colour. The L* value represents the lightness value. CMYK: the process colour inks used in commercial printing, which are: Cyan, Magenta, Yellow, and Black (Key). Colorimetry: measurement of the wavelength and the intensity of electromagnetic radiation in the visible region of the spectrum (Encyclopædia Britannica) ColorSync: colour management architecture that is built into Macintosh computers. Colourfulness: (see Saturation) Colour constancy: an effect of colour adjustment in human processing, is dependent on the brain processing and compensating for light changes in 8

9 the environment. For example, an environment over the course of the day will change depending on the amount of light, but our brain will adapt to changes of light levels. A camera that uses film can record these differences because it cannot take into account or compensate for the influence of light, for a digital camera the automatic white balance can compensate for differing lighting conditions. Colour measurement: a physical measurement of reflected light from an object under a known lighting condition. Colour space: a three dimensional model that encompasses all available colours. The shape of the space might be determined by the greatest extent of measured colours. A colour space can be defined by a physical collection of samples (e.g., Munsell), or by a mathematical system (e.g., CIELab). Colour temperature: is more subjective, specific colours may appear to be warm or cold, for example, red and orange suggest warmth and blue suggests cold. Within the hue colour range, vermilion appears to be warmer than rose madder, however, an ultramarine also appears warmer than a cobalt as the hue shift of the ultramarine tends towards a red hue whereas the cobalt tend towards the green. A useful example of the relationship of watercolour pigments is demonstrated by Bruce MacEvoy who has measured and plotted the colour locations on the CIECAM a C b C plane. Continuous Tone: a smooth continuous transition between light to dark tones with no visible halftones. Photographs and photomechanical processes such as photogravure, collotype, Woodburytype are examples of continuous tone images. Contrast of tone: different values of the same hue as opposed to contrast of colour, which is different hues of the same value. 9

10 Delta E ( E): a measurement for the total colour difference between two colours. Draw-down: the K-Bar, can be used for a wide range of coatings, has a micro diameter wire, which is wound precisely around the shaft. Different thicknesses of wire provide different deposits of ink. In this case, the K-Bar is dipped into the ink, the wire holds the ink is then held horizontally and pulled evenly over the paper, resulting in an even deposit of ink onto the substrate. Gamut: the set of all colours displayable or printable by a particular output device. Gloss: the angular selectivity of reflectance, involving surface reflected light, responsible for the degree to which the reflected highlights or images of objects may be seen as superimposed on the surface. (MN) Halftone: an image that is made up of a series of small squares and dots, which vary according to the intensity of the tone. The conventional method of halftone screening is one that uses a regular pattern of dots in straight rows or AM (amplitude modulation) screening. The dots are placed at a fixed frequency, with lesser or more dots corresponding to the required level of grey. This is also known as a clustered halftone dot. Hexachrome: an extension of the CMYK process colour to make a sixcolour set. The addition of red and green are used to expand the printer gamut. Hue: is loosely defined by generic colour terms, such as red, blue, yellow. Within these generic terms are particular colour names such as rose madder, vermillion, which can be both termed as having a red hue, in the same way that, ultramarine and cobalt have blue hues. A hue can also be described by its chemical element, such as cadmium red, or cobalt blue. HSB/HSV: Hue Saturation and Brightness/ Hue Saturation and Value 10

11 ICC or International Color Consortium: was established in 1993 by the major suppliers of operating systems, software and devices, to develop a standard for a colour management system (CMS). ICC Profile: describes the colour attributes of a particular device or viewing environment Illuminant D50: a reference to describe the colour temperature at 5000 degrees Kelvin. A standard lighting is used for evaluating printed colour on paper, in the graphic arts industry. Illuminant D65: a reference to describe the colour temperature at 6500 degrees Kelvin. This standard tends to be the default for monitor colour. LAB: CIEL*a*b* or colour space CIELAB is the space that describes all the colours visible to the human eye. The colour coordinates are plotted in a three-dimensional space, where the L* axis represents lightness and darkness, the a* represents the green to red axis and the b* represents the blue to yellow axis Opacity: the ability of a specimen to prevent the transmission of light. (MN) RGB Colour Space: as exampled in the red green and blue phosphors of a display device such as television and computer monitors. A colour in RGB space is represented by three numbers, which correspond to the signals of red, green, and blue. RGB colours are also called additive colours; where colours overlap, they produce cyan, magenta, yellow, and where all three are added they will produce white. RGB mode: 0 (black) to 255 (white) for each of the RGB components in a colour image. For example, a bright red colour might have the combined values of R-246, G-20, and B-50. If the values of all three components are 11

12 equal, the result is a shade of neutral grey. If the value of all components is 255, the result is white; and Black if all the values are at 0. RGB TC9.18: an industry standard colour chart, containing colour patches of known RGB colour values. The chart is designed to be measured using a spectrophotometer, which can then be used to create ICC profiles. Rastor Image Processor or RIP: in some printers such as the Hewlett Packard machines, a rastor image processor is integral to the hardware. Other printers require an external RIP, which is a software or hardware interface for processing and converting a digital image file for printing. The RIPs will contain a range of output profiles to obtain the maximum gamut for the capability of that printer. Saturation or chroma or colourfulness: of a colour describes in terms of how the saturation is reduced by adding grey (black and white), which will result in a degrading of a colour from intense (most saturated) to neutral (grey or least saturated). By mixing two or more hues does not reduce the saturation, ie. a purple (where red and blue is mixed) or brown (red, blue and yellow) where only hue colours are mixed (a grey is not added). This is different to comparing a hue on a digital colour wheel or spectrum where brown (as a hue) would be described as a dark orange or dark red (black would be added). Chroma is determined also by the brilliance or brightness of the colour, for example vermilion has a higher chroma than rose madder, however vermilion appears to be lighter than rose madder and so the tonal value of the colour is different. Simultaneous contrast: due to our lack of an absolute colour memory, it is impossible to avoid simultaneous contrast: colours are perceived relative to another. The context and ratio of one colour over a second colour will affect the way the second is perceived (figure 1). 12

13 Fig. 1 An example of simultaneous contrast Screenprint: A print process by which a fine mesh screen is coated with a photosensitive layer of emulsion and is exposed to UV light through a photopositive stencil. The screen is washed to remove the unexposed areas and then dried. Ink is then drawn through the screen to create the image on the paper surface. With modern meshes, much finer details and halftones can be obtained. Soft preview: Using colour management, a preview or proof of how the image will print can be viewed on the monitor. Spectrophotometer: a photometric device that measures the reflected light from an object. Measurements are made and referenced to known lighting sources (e.g. D65, D50) and Standard Observer conditions (e.g. 2 ). Spot Colour: a colour printed with a pre-mixed custom ink, rather than with a combination of the standard process colours. A spot colour can be described by CMYK percentages. Standard Observer: the CIE conducted experiments to quantify the perception of colour by a standard observer. In 1931 and 1964, experiments were undertaken based on a 2 field of view (similar to looking at one s thumbnail at arm s length), which resulted in establishing the 2 observer. In 1964, experiments were undertaken using a 10 field of view, which established the 10 observer standard. Stochastic: a process that uses a more random dot FM (frequency modulated) screening. Unlike the traditional halftone system, the process 13

14 employs no fixed grid or screen angles and the size of the dots and spacing between the dots is variable. Subtractive Colour Mixture: produced by combining colour pigments or dyes, using the primaries of cyan, magenta and yellow. By combining all three primary colours, the resulting image appears as a dark colour or ideally black (since all light is absorbed). Test chart: a series of colour patches used to generate ICC profiles for input and output devices. One of the frequently used test charts is the RGB TC9.18, and the ECI2002V CMYK, which is used for RGB and CMYK printers. Translucency: the property of a specimen by which it transmits light diffusely, without permitting a clear view of objects beyond the specimen and not in contact with it. (MN) Transparency: the property of a material by which objects may be clearly seen through a sheet of it. (MN) Value or tone: is a relative measurement from white to black or from lightest to darkest. The value of a colour can be described as to the extent to which a colour reflects or absorbs light. A colour by its own intrinsic value, ie. ultramarine blue, absorbs less light than cobalt blue, which absorbs even less light than Prussian blue. Wavelength: the distance between the crests of two adjacent waves crests, used to describe frequencies of light. 14

15 3. An overview of halftoning and digital print technologies The following sections provide an introduction to early halftoning methods, the development of stochastic halftoning screening and an overview of current inkjet technology. Continuous inkjet has been in use since the 1960s, and drop-on-demand from the 1970s. The last 15 years have witnessed the breakthrough into the desktop market and, more recently, the industrial and commercial products have developed to super wideformat and web-fed printing machines, where quantity is measured in metres per second. 3.1 Traditional Halftoning The conventional method of halftone screening is a method that employs a regular pattern of dots in straight rows or AM (amplitude modulation) screening. The dots are placed at a fixed frequency, with lesser or more dots corresponding to the required level of grey. This is also known as a clustered halftone dot. Since 1885 the ability to print in three colours was made possible by Frederick E. Ives. His technique employed the same principle as Maxwell s theory, of exposing an image through three coloured filters. Ives appropriated Fox Talbot s photogravure process of breaking down an image into small dots, but Ives utilised a more symmetrical matrix. Ives process was also based on Newton s theories, which Le Blon put into practice, using the three colour sensations: red, yellow and blue. For the purpose of a halftone, Ives devised a way of tricking the eye into reading an image that was made up of a series of small squares and dots, which varied according to the intensity of the tone. He made a series of opaque lines of equal width to their separation ruled on glass. Two rulings at right angles gave a pattern consisting of small openings on an opaque background. This process offered a greater depth of tone and made three or four colour printing more practical. A problem arose however when more than four colours were printed, giving rise to interference or Moiré patterns. 3.2 Stochastic Halftoning Stochastic technology can be considered as the beginning of a mechanical 15

16 simulation of 19th century continuous tone. The term stochastic is derived from the Greek word Stokastikos meaning to predict or guess and is used in mathematics to analyse and predict the movement of particles through liquid. In printing, it is a method that applies the placement of a more random dot FM (frequency modulated) screening. Unlike the traditional halftone system, the process employs no fixed grid or screen angles and the size of the dots and spacing between the dots is variable. Therefore more dots can be placed in a specific area in order to achieve a greater concentration of colour. Also, due to the removal of screen angles formerly preventing moiré effects, the stochastic system enables more than four-colours to be printed. Extra colours have been added to the process colour range, which include process orange and green. This has created a greater depth of colour, range of tone and improved contrast. Contemporary screening is hardware and RIP dependant. Wideformat inkjet machines and some smaller desk jet printers have inbuilt RIPS that convert digital files to inkjet near-continuous tone. PhotoShop grey-scale images can be converted to a crude bitmapped dither, which gives a useable stochastic quality. Fig. 2a Stochastic patterns Fig. 2b Conventional and stochastic halftoning 16

17 The previous six figures in 2a, are examples of different stochastic patterns, at the same enlargement (100%) of a 40% tone patch. The group of four figures in 2b illustrate a conventional halftone: on the top left, the spots form a symmetrical dot within a matrix, on the far right the stochastic spots are not constrained by matrix but appear to form a random pattern. The printed images underneath are corresponding magnified versions using offset lithography (left) and ink jet (right). 3.3 Basic types of digital print technologies Continuous inkjet describes a process in which minute drops of ink are generated continuously at around 100,000 drops per second. In order to not print in an area, there are two approaches: a binary system in which drops are printed or not printed (Kodak Versamark, Domino, Videojet); or as a Multi-level deflected system by which ink is deflected into a gutter, filtered and recycled (Domino, Imaje, Linx, Videojet). Drop on Demand describes a method in which drops are generated only when required. From a row of nozzles each contains a pump that fires a drop when needed, with dots generated at 10, 000 drops per second. There are three types of drop on demand approaches: Thermal or Bubble jet (Canon, HP, Fuji-Xerox, Lexmark, Olivetti); Piezo (Epson, Hitachi, Spectra, Xaar, Xerox); and Electrostatic (Seiko-Epson, TTP). Fig. 3 Inca Digital, The Columbia Turbo 17

18 Type DOD Thermal/ Bubble jet DOD Piezo DOD Piezo flat-bed DOD Piezo web Typical machine HP psc 1210 allin-one 60 Epson Stylus Photo R Inca Digital Columbia 350K Dotrix.factory 600K 870K Inks Speed Dpi Paper Size Vivera Up to A4 Aqueous ppm - colour dpi x inks normal - A CMYK+ (210 x 29) dpi Lc Lm DURA Brite Ultra pigment Ink system Aqueous pigment inks CMYK UV curable inks CMYK Agorix UV curable inks CMYK Up to 19 ppm - B/W - A4 (210 x 297 mm) Up to 18 ppm - colour - A4 (210 x 297 mm) 5760 x 1440 optimis ed dp A4 130m2/hour 800 dpi 3.2m x 1.6m up to 40mm thick 907 sqm per hour Table 1. Technology comparison: Drop on Demand (DOD) 2 300dpi + greysca le Max print width 63cm Print head HP Thermal Ink Jet printheads Epson Variablesized Droplet Technolog y for minimum 3pl droplet size Printhead of 180 nozzles for Black, SPECTRA piezo printheads Ttec greyscale printheads 2 Examples in tables 1 and 2 are adapted from a training session by Willis, M. and A. Hudd (2005). The Ink Jet Academy - The Theory of Inkjet. Digital Printing Conferences 2005, Geneva, Switzerland, IMI Europe. 18

19 Type Continuous inkjet Single nozzle/coding Continuous inkjet multi level heads Typical machine Domino A300 6,000 Creo Varis $30,000 Inks Speed Dpi Print head Aqueous/ 4m/sec 60dpi CIJ multideflection solvent inks (43m2/hour) heads Aqueous inks CMYK 6.1m2/hour 16 pages per hour 1,500d pi Creo CIJ multi-level heads Continuous inkjet web feed Versamar k 3700 $1-2 M Aqueous inks CMYK Table 2. Technology Comparison: Continuous Ink Jet (CIJ) 1-2,000 pages/min 6,100m2/hou r 300dpi Kodac CIJ binary heads 3.4 How the basic digital print processes work Fig. 4a Continuous Drops deflected into a gutter, filtered and recycled Drops generated continuously at 100, 000 drops per second 3 Fig. 4b Drop on demand Rows of nozzles, each with a pump to fire drops selectively Drops generated only when required at 10, 000 drops per second 3 Figures 4 and 5 are adapted from a training session by Willis, M. and A. Hudd (2005). The Ink Jet Academy - The Theory of Inkjet. Digital Printing Conferences 2005, Geneva, Switzerland, IMI Europe. 19

20 3.5 Drop on demand Drop on demand ink jet heads are now more affordable, with a low cost ink system that is compact. Typical drop volumes of individual ink drops are in the range of 1 to 5 pico litres, producing printed spot sizes on paper in the range of 10 to 20 micrometers in diameter. Its limitation is the speed by which it can make the drops (5,000 to 20,000 Hz per second) resulting in a slower process that is more suitable for home or office environment rather than for industrial applications such as web fed technology. 3.6 Industrial drop on demand Inkjet requires a series of compromises based on cost, speed, resolution and application. For industrial applications speed takes precedent over resolution, therefore as indicated in the technology comparison chart, speed for industrial printing is measured in metres per second/minute/hour compared to sheets per minute or hour. The opportunities for inkjet in industrial applications is increasing, and now overtaking the traditional litho, screenprint, gravure and flexographic market. These applications range from greetings cards and magazines to security documents such as identity and credit cards. Not only can ink be passed through the heads, any fluid including biomedical fluids or conductive materials for printed circuit boards have the potential to be printed. Inkjet is still a niche market, but this will change as more cross-over occurs between ink suppliers and hardware makers, and the cost of manufacture decreases. The capabilities of Piezo printhead hardware, with higher drop production rates and further ink development possible, suggest there is more scope for improving on resolution and speed. In comparison to thermal, the Piezo head can obtain a more controlled drop. However, any impurities in the ink can cause head failure and degeneration of the drops. 3.7 How a drop is formed in Piezo head technology The heater inside the reservoir heats up the ink, the heat is then suddenly removed resulting in very short pulses of 2-10µS. The temperature of the 20

21 heater rises to C resulting in a thin film of ink above the heater, which nucleates and vaporises forming tiny bubbles that then coalesce into one bubble. In a few microseconds the bubble expands rapidly, it then explodes causing a pressure wave within the channel and the ink is forced out of the nozzle. The heat is removed, the bubble collapses, the meniscus breaks the ink ligament and residue ink is drawn back into the channel. The process takes 10 microseconds to make the drop and 100 microseconds to return to its normal state. The following diagram shows a very simplified procedure indicating the level of pressure waves moving backwards and forwards inside the chamber. Meniscus at rest Meniscus disturbed by pressure wave Jet formation process initiated Surface tension creates droplet Fig. 5 How a droplet is formed Droplet formation within a Piezo ink chamber is similar to magnetising a magnet; the Piezo materials or crystals are electromagnetised. When an electrical charge is applied, the Piezo crystals deflect within the ink chamber (similar to a vibration of a drum skin), which creates a pressure pulse, thus ejecting ink from the nozzle. 3.8 A description of existing inkjet colour systems Since 2000, inkjet inks have been developed for a higher degree of permanence. The cheaper and less permanent inks fulfil a necessary role for major sectors of the industrial and office market, and the high end fine art market inks have significantly shifted from dye based to pigment based inks. 21

22 Pigment inks have brought longevity of printed output in line with standards established in the high-end museum quality print market. Although dye based inks present a more saturated colour gamut, it was necessary from our bias towards high quality fine art printing to move towards pigment based inks. Although pigment inks are less intense than dyes, they are more resistant to UV and gas fading. The pigment molecules are more complex and break down at a slower rate than the simpler, small molecule dyes. Whilst early pigment inks lacked intensity, advances have been made to improve the quality of these inks. This shift has developed in response to user demand for archiving and colour longevity, especially for the production of large print works, including exhibition, display and the fine art and poster market, where it is essential for colours not to fade. For the medium and wideformat market, the initial four CMYK ink sets have recently been expanded to eight and twelve colour sets. For example Canon s LUCIA Pigment Ink Technology contains twelve-colour inks (a mixture of pigment and dye colours) using CMYK and red, blue, green, grey, photo grey, photo cyan, photo magenta, and matte black. HP have introduced their Vivera range of colours to their new Design Jet Z series, which includes CMY plus light grey, grey, matte black and photo black, orange/red, blue, green, light magenta and gloss. By adding colours to the existing colour set to address gaps in the colour gamut, the inclusion of green has resulted in the light cyan being redundant. Epson s UltraChrome K3 eight-colour system has included two different black ink modes - photo black and matte black; Canon differentiates between matte black, regular black, grey and photo grey. According to Canon, the combination of both grey and photo grey enables smoother transitions from light to dark. HP have included four grey/blacks, with the addition of a gloss. As part of their media profiling management system this can be switched on or off to enhance density and is an automatic component of gloss papers 4. 4 The gloss enhancer is not available for matte papers. 22

23 Epson UltraChrome K3 ink incorporates a High-gloss Micro-crystal Encapsulation, in which, according to Epson literature, each pigment is coated in a resin to reduce the grouping of pigment particles. This is similar to Hewlett Packard s Vivera Electrosteric Encaplsulation Technology or EET; negative electrostatic charges are introduced within the resin layer, which coat the pigment particles and prevent them clumping together. 23

24 Chapter 4: A description of parallel experimentation into colour accuracy, papers and inks, and longevity Based on the initial results from The International Digital Miniature Print Portfolio, many areas of potential interest were highlighted as to possible problem areas for further research. It was necessary to separate these areas of interest, yet to allow the experimentation to be sufficiently dynamic that ideas formed during one experiment could be used to assist in a parallel enquiry. Many of the dominating issues related to colour still required further exploration. From an action research perspective, there was a fundamental requirement to develop a range of experiments in order to fully understand the process before this could be imparted to others. This knowledge required the inclusion of tools and methods for colour management, colour printing and colour appearance. With reference to the grounded theory, one important aspect of the methodology was the requirement for information to be understandable between different groups. I felt this to be vital in breaking down language barriers between the arts and science: complicated concepts could be simply explained and a description could replace a mathematical equation. The following sections of chapter 4 address how a range of enquiries were developed to inform a series of parallel research projects. The chapter will present experiments conducted in order to solve these problems, their results and the refinement of these ideas towards the development of alternative colours for inkjet printing. These early investigations sought to address a range of problems that were coming to the fore as the technology was evolving. Therefore the experiments were embarked upon with no definitive objective, but through a process of experimentation and testing that was based on addressing problems from the perspective of a fine art printmaker. From the benchmark of traditional methods of photomechanical printing using screenprinting inks 24

25 onto fine art papers, the experiments sought to address the three main areas of ink, machine and paper. 4.1 Accurate methods for printing colour During the three-day Wideformat Printing and Colour Master Class (July 2003), it was discussed with Senior Colour Scientist Johan Lammens of HP Barcelona, how artists could control colour more creatively, without requiring expensive profiling equipment. Lammens suggested a method using Photoshop that involved measuring just nine primary printed colour patches: cyan, magenta, yellow, red, green, blue, black (as achieved by overprinting cyan, yellow and magenta), white (the colour of the paper or ink) and black (ink). Based on this suggestion, I deduced that if I could measure any printed colour, I could work towards the creation of new profiles that would directly relate to the printed colours and how they would appear on screen in the Photoshop software. This was a very rudimentary approach to profiling, but in order to make an assessment between colours printed by an inkjet printer and other traditionally printed colours, then a comparison could be made as to the differences between digital and traditional colour printing methods. The objective was to identify the differences in the gamut, and then to begin to develop bespoke colours for an alternative ink-jet colour system that assisted in controlling how colour was printed. By working with the high quality capabilities of inkjet to create black-and-white half-toned separations which were based on measured screenprinted hue colours, decisions could be made more directly as to which colours were most useful. Screenprinted colour hues were considered to be the most appropriate, the inks were easily available, were consistent (colours were unlikely to vary from batch to batch), were related to the colour hues as utilised by many artists and had recognisable pigment terms. There were a wide range of colour hues, which included process cyan, magenta and yellow. The screenprinting process could render a smooth uniform layer of colour that could be measured and repeated with similar uniformity and accuracy. Colour separations could be accurately registered and printed, and colours could also be overprinted, thus enabling colour mixing. 25

26 4.2 Generating the colour circle In order to test the colour capabilities of a range of inkjet devices and to compare them with screenprint hues, a colour circle was generated in Photoshop. Hue Saturation Brightness (HSB) colour percentages were modified, and working in a CIE L*a*b* colour space; the a* component (green-red axis) and the b* component (blue-yellow axis). HSB was useful in this instance as a means of generating a 360 colour circle: by mathematically dividing the circle into equal segments and modifying the hues using saturation and lightness to obtain black and white. Two colour circles were generated (figure 6): twenty-four segments of equal degrees form the outer ring of the colour wheel. Two colour wheels were devised: 100% hue colour to white and 100% hue colour towards black. Fig.6 Two colour circles generated in Hue Saturation and Brightness. For each circle, the segments were divided into 10% increments, the first working from 100% (saturation and brightness) using the saturation slider to add white by 10% increments to 10% near white; then from 100% (saturation and brightness), using the brightness slider to make of 10% increments to near black (figure 7). 26

27 Fig. 7 HSB slider in Photoshop Inkjet printing the colour circles The inkjet tests were undertaken using the pigment based HP Z and the dye based HP 130nr 6 inkjet printers. As the Z3100 printer also had an extended inkset, which included red, green, blue, and grey, it was useful to compare the printable gamut between these two machines. The circles were printed onto A3 sized sheets onto Hahnemühle PhotoRag fine art paper. The circles were printed from the LAB colour mode, using application managed colour in Photoshop, and the settings best and heavyweight coated fine art paper, were used in the printer driver. Based on initial trials during the Digital Portfolio, we found that just by changing the output profile, resulted in very different colour prints. Therefore, I was also interested in investigating further how the colour information was managed and sent to the printer and, working on the Z3100, two output profiles were used and compared. In the first case, it was decided to keep the same application managed colour, and in the second a paper profile was selected to match the heavyweight coated fine art paper in the printer driver. 5 The inkset in the HP Z3100 contains: cyan, magenta, yellow, matte black, light magenta, red, green, blue, light grey, medium grey, (photo black, gloss these would not be used in this print mode). 6 The inkset in the HP 310nr contains: cyan, magenta, yellow, black, light cyan, light magenta 27

28 4.2.2 Preparing and making the colour separations for screenprint Unmixed hue screenprint inks of: process cyan, process magenta, process yellow, cadmium yellow, cadmium red, cadmium scarlet, ultramarine, leaf green, phthalocyanine green, black, purple (figure 8) were screenprinted onto a bright white card. Once dry they were measured using a Gretag Macbeth spectrophotometer, using an illumination D50 and at 2. Both L*a*b* values and spectral data were recorded. The measurements of each hue can be used to compare colour differences, create colour profiles and generate multi-colour separations in preparation for printing. Fig. 8 Primary hue inks screenprinted onto white card. Top left: cyan, ultramarine, cadmium red, process magenta, process yellow, purple. Bottom left: phthalocyanine green, black, cadmium scarlet, cadmium yellow, leaf green. The Eye-One Share software 7, which is the software interface for the spectrophotometer, enabled colour patches to be visualised in relation to other colours. The measured data was compared and exported, from which colour palettes could be generated, or as spectral charts as demonstrated below (figure 9). 7 The Eye One spectrophotometer and visualizing software from GretagMacbeth is a hand held measuring device that can measure spectral information from emissive sources such as a CRT/LCD display or reflective sources, such as print, paint or fabric samples 28

29 process cyan ultramarine cadmium yellow cadmium red phthalo green leaf green process yellow process black cadmium scarlet process magenta purple nm 400nm 420nm 440nm 460nm 480nm 500nm 520nm 540nm 560nm 580nm 600nm 620nm 640nm 660nm 680nm 700nm 720nm Fig. 9 The spectral data of the screenprinted colours Fig. 10 L*a*b* colour measurements, from the screenprinted ink patches, are added to generate a custom CMYK profile. Following Lammen s detailed technical notes on how to create profiles in Photoshop 8, the measured screenprint hues were added to Ink Colors (figure 8 Based on laboratory notes written by Johan Lammens and Pau Soler of HP Labs, Barcelona, 6 th March 1999, Quick ICC profiling with Photoshop 5 29

30 10) in Photoshop, and a new custom CMYK profile was generated. The following settings were used: Open File Colour Settings CMYK Setup select Built-in CMYK model Select Ink Colors Custom, check the LAB coordinates box, uncheck Estimate Overprints box and input the 9 values in LAB (Figure 2). For the section on dot gain, Lammens suggests 15% for glossy papers and 20% for coated papers. For separation options he suggests Separation type: GCR; Black Generation: Medium; Black Ink Limit: 100%; Total ink Limit 240% for glossy bi-level, 200% for native coated bilevel 400% if the ink limits are applied through the printer; UCA amount 0; save as xxxx s new profile. This profile was named Screenprint Colours. This profile could then be used to preview how the colour circle would appear when screenprinted and the potential limits in the screenprint colour gamut. In order to make a multi-colour stochastic separation suitable for screenprinting, the LAB colour file had to be converted to a CMYK file. It was found that when converting from LAB to a generic CMYK space, there was a significant reduction of the colour gamut resulting in a de-saturation in colour. Two alternative CMYK profiles were tried, which included the new Screenprint Colours profile and a commercial HP profile for a fine art coated paper. On initial inspection from the monitor screen, the HP profile demonstrated the least colour difference and better information in the green areas, whilst the Screenprint Colours profile, which was based on just nine measured hues, offered a good alternative, especially in the reds, magenta and orange region. This simple exercise demonstrated that there was not one simple approach to colour conversion. The Screenprint Colours profile was used to convert the colours from LAB to CMYK mode. PhotoSpot, by Second Glance Software Inc., a colour separation plug-in to Photoshop, is a useful software for working with bespoke colours and separating the image according to these colours. It was important to bear in mind that the conversion from an RGB or LAB colour mode to CMYK mode would result in a reduced coloured image. Therefore it was useful to convert the image using the custom Screenprint Colours 30

31 profile. Using PhotoSpot, two different colour separations were generated, along with an eight-colour and a nine-colour separation. The colours used for the two screenprint versions were: Screenprint (1): cyan, magenta, yellow, cadmium red, cadmium scarlet, ultramarine, leaf green, black. Screenprint (2): cyan, magenta, yellow, cadmium scarlet, ultramarine, leaf green, phthalo green, black. Fig. 11 The original HSB circles in LAB mode, in comparison to the same circles converted to a generic CMYK colour mode. The circles on the left are the unconverted colours in photoshop and the circles on the right demonstrate how the colour values appear when converted to CMYK colour mode. 31

32 Fig. 12 These two circles are converted using different CMYK profiles, the first is based on a coated fine art paper, which offers the largest CMYK gamut and compared to the 11 measured values from the screenprint hues Results The following figures (14a-e) illustrate measured L*a*b* data from the twenty-four hue patches. These appear on the outside perimeter of each colour circle. Figure 13 shows a colour circle that has been rotated so that the yellow appears at the top and has been flipped horizontally. This is so that a direct comparison can be made with the measured data, the graphs, and the standard CIE L*a*b* colour circle: yellow at the top, blue at the bottom, red to the right and green to the left. 32

33 Fig. 13 Colour circle rotated and flipped to the a* b* axis of the measured charts. Fig. 14a Screenprint hues Fig.14b HP Designjet 130nr. 120 Fig. 14c Screenprinted colours from colour separation Fig. 14d Z3100 using an output paper profile. Fig. 14e Z3100 output same as source. 33

34 yellow Screen-ink hues green red Screenprint with colour separation HP 310 HP Z3100 HP Z3100 with profile blue Fig. 15 Comparing two profiles that were used to convert to a CMYK colour mode. Note the loss of saturation in the green section of the circle on the right. By over-layering each plot in figures through 14a to 14e, a comparison can be made, demonstrating a larger gamut of the screen ink hues, as represented by the turquoise line (figure 15). The curve of the turquoise line around the bottom of the circle demonstrates more blue and purplish blues in the screen ink gamut, with the Z3100 showing more of an inclination towards the greens, due to the inclusion of the green cartridge Conclusion There was a significant difference between the two circles printed on the Z3100 using different output profiles. Even though the source image was the same, by changing the output profile, one could dramatically change the outcomes of the printed colour. Both circles utilised the application driven colour setting by driving colour data from the application instead of using the printer default software. The addition of the extra red and green inks in the Z3100 demonstrated an advantage over the 130nr six-colour printer and a larger printable colour gamut. 34

35 The measured screenprinted hues (figure 15) showed the largest gamut of all, especially in the blue region. Even though the same screen inks were used to print the circle, if compared to the screen ink hue colour patches, the screenprinted circle resulted in a more reduced colour gamut. As demonstrated in the figures 11 and 12, by converting from LAB to CMYK, there was a noticeable colour reduction in different sections of the colour wheel, especially in the green areas. This could be improved by using a larger gamut CMYK conversion space, such as a custom designed paper profile. The findings from the experiments demonstrated that the hue screenprint inks, when measured independently, revealed a much larger gamut than the same inks used for screenprinting the colour separations. However, during the process of colour conversion the potential of the colours was much reduced. The same problem could be hypothesised for the inkjet colours that, at the point of converting from one colour mode to another, or from image file to printer, there is an incremental loss of colour information. If for example, a more direct approach to digital colour mixing that was less reliant on colour management could be found, there could there be a greater control over each of the mixing and printing of these colours. 4.3 Inkjet papers and inks During the 90s, hardware for capturing and printing digital images was becoming more widely available. However, the problem arose when printing images onto standard inkjet papers; artists had a much more reduced range of papers than they were accustomed to using in the field of fine art. These standard inkjet papers tended to be lightweight, bright white, low-grade wood pulp proofing papers or plastic backed glossy photo-style papers. The paper specification typically included a paper surface that emulated a photographic white, with smooth, gloss or matt finish that provided optimum conditions for 35

36 printed colour reproduction, but not necessarily the primary requirement for the artist 9. For traditional printmaking processes, the artist has a wide variety of fine art papers from which to choose. The quality of the paper is determined by the manufacturing processes and the content of the paper-pulp material. Many of these fine art papers are of a high quality that is either hand-made or mould-made. They typically contain cellulose fibres from the cotton plant. They are archivally stable and the fibres, in general, do not break or yellow with time. The papers are generally classified by weight, ranging from lightweight 12gsm (Gampi Japanese tissue) to a heavy-weight paper of around 350/400gsm (Du Chene - a hand-made paper specially made at Moulin de Larroque). The fine art papers tend to have a surface texture and a natural off white colour (white, off-white and cream). In 1999, there were very few papers that could have been considered by artists to be suitable for fine art printing. It was also quickly recognised that many of the commercial low-grade papers could contribute to colour fading, which was also an important consideration if an artwork was to be displayed and sold. In order to address the limited availability of suitable papers for artists, and to ascertain whether uncoated fine art papers could be used, a series of print tests were undertaken. The project began in consultation with John Purcell Paper and St Cuthberts Mill, makers of the Somerset Enhanced paper, 10 and an investigation into coated and un-coated printmaking papers was undertaken. Uncoated paper can range from commercial office photocopier paper to fine art printmaking paper. The quality of the paper is dependent of the grade of 9 A common assumption is that printed images will be framed or presented flat. For the artist wishing to adapt prints into three-dimensional works the paper is an important consideration, for example, if the print is cut or folded to be made into a book or a sculptural work, an inkjet paper would require coating on both sides and could withstand scoring, creasing and folding. 10 In 1999, St Cuthberts Mill had responded to the artist market with the launch of the Somerset Enhanced paper, which incorporated an inkjet receiver layer on the surface of an archival paper; and later with the Bockingford Inkjet (2002), which is coated on both sides. 36

37 cellulose fibre used in its manufacture. For fine art papers, only the best materials are used, such as linen or cotton. For low-grade papers, a wood pulp is often used. A coated paper is an inkjet paper that has a thin receiver layer coating to retain the ink on the surface of the paper. Emulating photographic terms, these coated papers are generally classified as matt, gloss and semi gloss. The three optimum qualities for a coated paper are smoothness, whiteness and brightness. Based on his extensive knowledge of fine art papers, John Purcell supplied a preliminary list of papers, which he considered had potential surface qualities for inkjet printing that were popular with artists. As described by Richard Shearwood-Porter of St Cuthberts Paper Mill, the ideal qualities of a printed image are good colour separation, good colour lift, tone proportional to ink volume, true black and a high level of definition. 11 These qualities were agreed to include and related to the paper in terms of: smoothness a paper needed to have a uniform surface so that the ink heads would not catch on the surface and the ink could be applied evenly; weight an optimum weight was crucial when handling paper, especially if the sheets were large, to ensure the paper did not crease because it was too thin, and not so heavy that it could not be fed through the printer; whiteness the body colour of the paper could assist in maintaining bright colours; sizing the sizing of the paper might help in preventing the ink from spreading or sinking into the paper, and keep the edges of text sharp. 11 Addressed to an audience at the Preservation and Conservation Issues Related to Digital Printing and Digital Photography Conference, Institute of Physics, London, 26 th -27 th October

38 At this time, it was considered that the methods for the creation of a digital print were frustrated by the technology. Firstly, not enough was understood as to the relationship between what was visible on screen and how accurately the digital information could be transferred from one hardware device to the next. Secondly, it seemed that there were many possibilities that colour data could be mistranslated in the colour-printing pipeline. The opportunity by the artist to understand how the problems were arising, were obstructed by the limits in colour management systems. The objective was to test the process by comparing the visual information and grading the printed image according to the quality of the original to the printed reproduction. These assessments were entirely subjective and non scientific, and were based on a criteria of what was most pleasing to myself as a printmaker and an evaluation of the relative qualities of ink and paper according to traditional printmaking practice: the texture of the paper, the impact of ink on paper; the image quality of colour and tone, and the density of text Early pilot study with Encad Novajet printer In 1999, a pilot study using the Encad Novajet 63E 600DPI printer was undertaken. A series of print tests were considered, which looked at a variety of artist's archival papers, and fine art inkjet coated papers, and took into account subjective responses to the effects of ink on paper. Following an approach of comparing an original to a reproduction, sections of an artwork were scanned and imported into Photoshop. These included traditional print processes from a screenprint, intaglio, collograph, monoprint, commercial printed maps that contained fine detailed lines and specific colours, and found textures such as a gloss surface, white muslin and a blind embossed surface (figure 16). In Photoshop a grey-scale, a colour-scale, and text was added. The final digital image was then printed onto a range of papers, which were then compared to the original artwork. 38

39 The papers that were tested ranged from lightweight Japanese tissues to heavyweight etching papers. There was also a variation in the base colour of the paper, which ranged from a bright white to an antique cream. Paper and Ink Trials at the Centre for Fine Print Research, UWE Onto a range of artist s printmaking papers on the Encad Novajet pro 600e 42 /1067mm including Somerset VelvetTP Satin Textured Enhanced Hahnemühle BFK Rives Arches 88 Velin Fabriano Tiepolo Artistico Rivoli Rosapina Moulin de Gue VAB Fig. 16 The first test page composed for the Encad Novajet, showing artwork, line work, greyscale and colour scale. As there was not a dedicated profile available for the paper and ink combination, one profile was chosen and was used for all the papers. A series of iterative tests on each paper were undertaken until the closest colour match could be obtained. Only by adjusting the percentages of CMYK inks in the printer server hardware was it possible to obtain the right colour 39

40 balance 12. Even so, when compared to the original artwork, it was impossible to obtain the richness of some of the colours and render an accurate texture Results of the Novajet experiment The results, within the parameters of the experiment, displayed a broad range of potentials, but further exploration was required. It was considered that the main problem of image reproduction was the global impact of colour correction. By adding more cyan to one area, the brightness of the green or blues might have improved, but could also de-saturate or darken another colour in a different area. Paper Colour Ink on paper Text Arches 88 No movement (2) Somerset white Satin (8) And Waterford HP (2) Heritage white (2) Arches Aquarelle Satine (2) Somerset White Velvet (4) Appears grey in the dark areas, loss of midtones, grainy, Noise in colour Slightly darker. Loss of detail in dark areas. Loss of detail in dark areas. Loss of detail in light areas. Slight loss of detail in dark areas Slight loss of detail in dark areas Over saturation of ink resulting in bleeding from black into the red square, also dragging and feathering along the edge of the brown square Ink bleeding from black into red Ink bleeding from black into red No movement Ink bleeding from black into red, also dragging and feathering along the edge of the brown square Holds edges, no feathering Slight feathering Black appears grey Ink patchy and uneven Black appears grey Ink more patchy and uneven than Waterford HP Holds edges, no feathering Good tonal line Holds edges, no feathering Good tonal line 12 The number in brackets in the results section refers to the number of attempts to obtain a close perceptual match. 40

41 Washi (5) For closest match, loss of detail in the lightest and the darkest areas Very slight loss of Ink bleeding from black into red Holds edges, no feathering Good tonal line Rayon Slight feathering at Good tonal line detail the blackest point Text is black Image Pro Good match No movement Holds edges, no feathering Good tonal line Somerset Enhanced Washi for inkjet Slight loss of detail in dark areas Slight loss of detail in dark areas Table 3. Results of the first Novajet experiment Slight feathering at the blackest point No movement Holds edges, no feathering Good tonal line Holds edges, no feathering Good tonal line Conclusions The Novajet commercial inkjet paper demonstrated the closest colour match to the original artwork. This firstly confirmed that the paper profile was suitable for this paper and ink combination, and that there was an accurate colour workflow from the scanned image to printed ink on paper. Secondly, the rendition of ink on the art papers highlighted the different results on the variety of paper from the same outputted image using the same paper profile. The criteria for assessing the paper were based on a comparison between the original artwork and how the ink was absorbed into the fibres of the paper. In some cases, the Waterford and Heritage paper repelled the ink resulting in a patchy surface, similar to the dimpled appearance of orange peel. Other papers, where the colour may have been absorbed into the surface, showed a rich and dense quality but had a loss of detail in the dark areas. The Japanese papers, capable traditionally of withstanding the application of watercolour paint, demonstrated that they could also hold a volume of inkjet ink with little movement in the ink. 41

42 4.3.4 Ink on paper trials using the HP5500ps In response to improvements in hardware and software, the addition of colours to the inkset and improvements in resolution and pigmented ink, new paper trials were undertaken ( ). 13 The HP5500ps six-colour printer, with the addition of light cyan and light magenta to the inkset (CMYKcm) was used to print onto coated and uncoated papers. This printer contained a pigment inkset, with a resolution much finer than the Encad. A new template was prepared in Adobe InDesign. The images in figure 17 were made by (starting at the top left hand corner and moving clockwise) Paul Thirkell, Carinna Parraman, Paul Laidler and Steve Hoskins. The CMYK round target has been generated at 100% cyan, yellow, magenta and black in Adobe Photoshop and saved as a tiff file. The CMYK target for the Epson RGB has been generated using 255 RGB (red, green, blue) and 0 RGB (black). The photographs of the flowers were taken with a Kodak 2.2 mega pixel digital camera. The JPEG images were opened in Adobe RGB ( 98) colour space and saved as a tiff. The grey scale photograph was scanned on a Linoscan 1200 in LAB, transferred to Photoshop, converted to greyscale and saved as a tiff; a second layer, which is an Illustrator file, was laid over the image. The kite was generated in Freehand and then imported to Photoshop, with a second layer added and saved as a.psd file. In order to include an expanded range of papers that might be suitable for the artist, a much greater range of uncoated papers was chosen. 14 St Cuthberts Mill also requested ink on paper tests of their prototype range of new generation Somerset Enhanced and Bockingford Inkjet papers, and the same template was used for these papers. After testing a range of paper profiles on a sample range of papers, the optimum output profile and paper profile was selected and used for all the papers. The selected paper profile 13 The project began by using an Epson and HP printer. Based on a hypothesis that papers might react to piezo and thermal printers in a different way, however after duplicating the tests on both printers, it quickly became apparent that the papers accepted the ink in very much the same way irrespective of printer. The HP5500ps was a much faster printer than the Epson, and it was decided to continue testing with the HP machine. 14 In consultation with John Purcell of John Purcell Paper, 42

43 was a heavy-weight fine art paper and the best setting was used in the printer driver software. Fig. 17 Test sheet for the HP 5000ps, comprising digital images and text. 43

44 Fig. 18 All the prints were arranged on a table and compared based on a series of criteria Ink on paper trials using the HP 5500ps ( ) initial results Once all the prints were made, all the prints were laid out on a table and a visual comparison was made under indoor lighting conditions with a north facing window, during the middle of the day, and no artificial lighting (figure 18). After an initial quality inspection, some papers deemed unsuitable were rejected; for example, papers that repelled the ink causing the image to appear patchy eg. Simili Japon, Fabriano 5, Bockingford, Rivoli (table 4). Paper Velin Arches Blanc 200gsm Aquarella Arches Rough 185gsm Arches 88 Aquarella Arches NOT 185gsm Colour appearance Good colour balance And tonal range Loss in mid-tones. Good black. Good black, and tonal range Slight loss in midtones Ink on paper Good edges Very slight movement in target edges Very slight movement in target edges Good Very slight Text Good Good Good Good 44

45 BFK Rives 250gsm Bockingford 150gsm 76x56cm Good black, information in light areas. Slightly grainy in the target shadow Slightly pale appearance, slightly grainy. Good black on the colour patches. Black appears grey in the colour patches movement in target edges No movement in the target Slight movement. More of a patchy appearance. Good Holds edges Black appears slightly grey Fabriano Artistico 50x70cm Fabriano 5 Hot Press 210gsm 50x70cm Fabriano 5 NOT 210gsm Hahnemühle Old Antique aid 180gsm JPP Mouldmade 185gsm 57x76cm Moulin Du Gue 270gsm Rivoli White 240gsm 70 x100cm Saunders Waterford Hot Press 300gsm Saunders Waterford NOT 190 gsm Saunders Waterford Hot Press 190 gsm Good black, and tonal range Same as NOT Slightly pale appearance Slightly pale appearance (Yellow paper, textured) Information in the light areas. Grainy appearance Colours appear slightly more saturated Slight loss of mid tones. Good tonal range Good black in the colour patches, slight loss of mid tones, information in light and dark areas. Good black in the colour patches, slight loss of mid tones, information in light and dark areas. Very slight movement on target edges Very slight movement on target edges, also patchy Very slight movement on target edges Holds the ink well Slight patchy appearance. Very slight movement on target edges Very slight movement on target edges Slight movement. Patchy appearance. Patchy appearance, Quite a lot of movement on target edges Slight movement. Slightly, patchy appearance. Slight movement. Slightly, patchy appearance. Good Good Good Good Slight movement Good Slight movement Slight movement Good edges Good edges 45

46 Saunders Waterford Rough 190 gsm Simili Japon 130gsm 96 x 64cm Somerset white Satin 250gsm Somerset Velvet Radiant White 250gsm Somerset Tubsized Satin 310gsm Somerset Text Laid 120gsm Somerset Satin 250gsm Zerkall smooth 100gsm Zerkall Ingres Laid 90gsm 48x65cm Good black in the colour patches, slight loss of mid tones, information in light and dark areas. Black appears grey in the colour patches Slightly pale appearance. No strong black Slightly pale appearance Slightly grainy Stronger black Slight loss of mid tones. Information in dark areas Slightly pale in appearance. Slight loss of mid tones, information in light areas. Slight loss of mid tones, information in light areas. Japanese Papers White paper. Bright clean colour and tonal range Tosowashi Good Torinoko Sekishu Slight colour loss in the highlights. Good black in the colour patches Bright clean colour, slight loss in the highlights Slight movement. Slightly, patchy appearance. More of a patchy appearance. Movement on the target edges. Very slight patchy appearance, otherwise sharp edges Very slight patchy appearance, otherwise sharp edges Slight movement on target edge between black and yellow Slight patchy appearance. Very slight movement on target edges Slight patchy appearance. Very slight movement on target edges Slight movement. Slight movement. Patchy appearance. Ridges of the papers interfere with the viewing of the image. Little movement Printhead scuff marks on paper, otherwise no movement Slight movement Little movement Table 4. Ink on paper trials using the HP 5500ps ( ), initial results Good edges Slight movement Good Good Good Good Good Good edges Good edges Good edges, text appears heavier Holds edges, no feathering Good edges Good edges 46

47 4.3.6 Ink on paper trials, results after refinement The remaining papers were compared and rated according to particular qualities in different printed sections of image (figure 19). Specific qualities included: sharpness of an edge between one single colour and another (coloured target) brightness of a particular colour (coloured target) tonal rendering of a colour (flowers) solid rendering of a light single colour (green kite) black and white tonal rendering (sea and sand) The results to the refinement process are recorded in table 5. Colour brightness Bleeding between yellow and black Tonal rendering in coloured Areas Tonal Rendering Solid colour Darkest point Fig. 19 particular areas that were compared during the refinement of the selection process. 47

48 Paper Green kite: a single light colour Sea and Sand: B&W tonal range Target: Bleeding between yellow and black Velin Arches Blanc Aquarella Arches Rough Arches 88 Aquarella Arches NOT BFK Rives Fabriano Artistico Hahnemühle Old Antique laid JPP Mouldmade Moulin Du Gue Saunders Waterford Hot Press Saunders Waterford NOT Saunders Waterford Hot Press Saunders Waterford Rough Flower: Tonal rendering of colour Somerset white Satin Somerset Velvet Radiant White Somerset Text Laid Somerset Satin Zerkall smooth Zerkall Ingres Laid Japanese Papers Tosowashi Torinoko Sekishu Table 5. Ink on paper trials, results after refinement Target: brightness of the blue Ink on paper trials, conclusions The papers were graded with ticks according to one 1 tick good, 2 ticks very good, 3 ticks excellent. The results demonstrated that none of the papers demonstrated any exceptional qualities. The papers with the most amount of ticks suggested the greatest potential for further experimentation. 48

49 These are: BFK Rives, Moulin Du Gue, Arches 88, Saunders Waterford Rough and NOT, Fabriano Artistico, Velin Arches Blanc, and the Japanese paper The Hahnemühle Old Antique laid paper, although the body colour of the paper has a yellowish quality, there are other white Hahnemühle papers that would benefit from further investigation. This yellow paper might be appropriate for printing an image that has an aged quality. Whilst the uncoated printmaking papers showed disappointing results, the combination of ink on paper nevertheless suggested a potential for improvement. If combined with other processes such as laser cutting and screenprinting, or fabricated into three-dimensional prints or artists books, then the requirement for perfect colour reproduction is less important. Furthermore, when using uncoated papers images can be printed on the front and back with no change in the quality of the image. From personal experience, when experimenting with inkjet printing onto uncoated papers that if subjected to embossing, cutting, folding, gluing and pressing, there was a greater possibility for a more robust print. It was quickly found that a coated paper would be too fragile for such an activity. The surface was easily damaged by water, scuffing, dirt and grease, and resulted in a print that was irreparable. It was concluded that both coated and uncoated papers were required. The coated papers offered the best colour, tone and density but little could be done to manipulate the surface once printed. Whilst the uncoated papers demonstrated a reduction in brightness and density, these papers could be subjected to more rigorous post-processing. A recommendation for the future of coated papers would be to include a gloss or a matt varnish coating, that would only be printed in selected areas of the print to enhance a an area or where there was printed ink on paper as a protective coating. As inkjet hardware and inks improve and new papers 49

50 are introduced, these experiments are regularly repeated, and an remain ongoing experiment. 4.4 The longevity of papers and inks In 2001, from discussions with curators at the Tate and with other artists and collectors during fine art print conferences such as the Southern Graphics Council 15 conferences and IMPACT conferences, it was discovered that print collectors were resistant to purchasing inkjet prints, which was mainly due to anecdotal accounts of the fugitive nature of Iris prints from the 1980s. Furthermore, as early inkjet printers continued to use dye-based inks, the combination of dye inks on artists coated paper exacerbated a continued reluctance from print buyers. It is less the problem now, but at the time when our research was undertaken, we found it to be a significant issue, that whilst the inkjet technology was capable of printing full colour images, if these prints faded after a short period of time then their visual qualities and value could be vastly decreased. The inkjet papers and inks trials also indicated there was a less of an occurrence of fading when printed onto uncoated papers. It was obvious that the ink was fading on coated papers; however the question was what was occurring to the inks on a micro scale? Indicators such as different ink combinations, how the ink sat on the paper surface of coated samples, or how the ink was absorbed into the fibres of uncoated papers, might be better understood if microscopy was employed and could assist in answering what was happening over time and further indicate what might be a good paper and ink combination. In 2001, with this in mind and in preparation for printing The International Digital Miniature Print Portfolio, an early investigation ensuring that the fine art prints made at the CFPR would last began. Working with Cranfield Colours, a series of accelerated exposure tests were initially undertaken on a small range of enhanced or coated papers. As there is no empirical data of how paper and ink combinations will fare over a long time, accelerated 15 The SGC conference is held every year at different host institutes in the USA. The impact conference at which originated at the centre fine print research in 1999, occurs every four years in different countries. 50

51 lighting methods are used to reduce the time and provide an indication of longevity. The collaborative research undertaken, that included KTP projects and work with HP, enabled the Centre to work outside fine art and design, entering the field of science and technology, for which we were able to purchase new testing equipment. This dialogue of discovery meant that as more knowledge was gained, more information needed to be gathered. As a result of working with Cranfield Colours on the development of inks, we had shared use of a Xenon accelerated lightfast test machine. Accelerated exposure is a process that increases the exposure of light intensity whilst proportionately reducing the exposure time. Many printer and paper manufacturers quote the WIR (Wilhelm Imaging Research, Inc.), which is a Display Permanence Rating, and is used as a comparison against other papers and inks. However, the research found that the problem with accelerated fading is that it could not take into account all the parameters that a print might be subjected to. Theoretically, accelerating fading by, for example, doubling light intensity levels and halving exposure times (as one would do with a camera: shutter speed versus light input) can give an approximation, but this law of reciprocity (calculating image degradation by simulating long-term exposure to light) is problematic and is only valid for particular conditions. Due to reciprocity failure, any anomalies can have an impact on the life expectancy of a print. The prediction can vary widely and is generally acknowledged to not be a reliable method in all circumstances. (B.Vogt, 2001, p.27). This was a very new area and an evolving problem for both the CFPR and printer and paper manufacturers, and for my research in obtaining the most appropriate and useful method for testing. My primary source of information was from the Wilhelm Institute (USA), which positioned itself early in the debate as an independent testing facility for lightfast testing and had provided some useful data over the previous decades. Better placed in the UK, other organizations such as the Institute of Physics and the London College of Printing had also highlighted their concerns through ongoing debate at the Preservation and Conservation Issues Related to Digital Printing and Digital Photography conferences (2001-). Based on my early 51

52 tests into the lightfastness of coated papers for inkjet, I concurred there was an increasing need to obtain useful and relevant data as to how these newly developed papers and inks would fare over time. Since the beginning of these tests in 2001, the quality of inks and papers has vastly improved, as has the print technology. The three main wideformat manufacturers in the UK began to ship printers with inksets that contained pigment inks, so the early problem of a trade-off between brightness and longevity had been largely overcome. Printer manufacturers, working within Wilhelm testing parameters, also guaranteed print longevity of more than 100 years. This was good news if the print had been mounted and framed to archival standards. However, as wideformat printers were becoming larger and computers were able to process large image files, there was more opportunity to make very big prints that moved beyond the boundaries of the frame such as free hanging works and installation works 16. It is very likely that these works could be subjected to a similar environmental impact as snapshot photographs attached to the refrigerator door, or pinned to walls in offices and homes and subjected to ambient air (Wilhelm, 2007). A test by Wilhelm established that the compatibility of paper and ink had a significant impact on the life expectancy of the image. For example, if an HP ink, paper and printer were used there was hardly any visible difference over 25 years. However, using a different paper with HP inks and printer resulted initially in a reduced gamut image and image degeneration in less than 3 months An assessment of appropriate methods for testing The problem was how to undertake an appropriate lightfast test, and the type of information that was useful to the artist in a real world situation in which light, humidity and temperature changes significantly over the year. As time and research into paper and ink progressed, the research highlighted that many parameters required consideration, which might have been beyond the initial scope of traditional lightfast testing. Artists working at the CFPR were 16 Particular examples were the room-sized pieces in Wolfgang Tilman s exhibition, If one thing matters, everything matters (6th June- 14th September 2003), at the Tate Britain, which demonstrated that as a room-sized work, the prints were intended only for the lifetime of the exhibition. 52

53 engaged in a range of digital printing and paper construction. This included large scale printing, and might be combined with other materials, such as stitching, screen inks, other printmaking processes, embossing and the surface of the paper might be dampened. These artworks might be stored in portfolios, in drawers, dry-mounted and unframed. Testing can be separated into predictive and comparative methods 17. A predictive test can attempt to make a suggestion on how long a print will last. Comparative testing demonstrates how printed samples will fade or change under particular conditions. Comparisons can be made, for example by measuring colour changes or paper samples, a visual inspection and a description of these changes, or through photography. Comparative testing does not predict the longevity of a sample. My interest in the relationship of ink and paper was based on comparative testing methods and gaining an understanding of what was occurring to ink and paper that was not accelerated. Although this method could be considered a subjective approach to testing, it provided useful information for users, giving an overview of changes over time, and enriched the research context of issues relating to lightfastness. The second part of the testing involved the measurement and recording of the changes that occurred to the ink. The two most common approaches to measuring are the blue wool scale - a method of comparing eight blue dyed strips of wool, that have different lightfastness properties, 1 being the most sensitive and 8 being the most resistant to fading; and Delta E ( E) difference - a measurement to describe the total colour difference between two colours. Whilst measurements are useful as a metric guide, a visual comparison is also useful. Therefore, the use of a microscope became both a tool for studying the relationship of ink to the paper surface, and recording the changes that were taking place to the ink dots over time Accelerated Tests for Fading and yellowish Stain Formation in Color Prints and Films Chapter 2, Page 64 53

54 4.4.2 Measuring colour changes during exposure to light under magnification The first experiment was an investigation into how inkjet colour was affected, firstly by the coated or uncoated paper, secondly exposure to light and thirdly, the combination of both. The research tested both pigmented and dye based inks printed onto coated and uncoated fine art papers, and were subjected to light conditions that were constantly changing throughout the day and consequently the year. 18 Colour patches were printed onto a variety of artist s coated and uncoated papers and then measured and photographed. The microphotography demonstrated how ink sat within the fibres of the paper, and how colour and colour combinations changed over time. Fig. 20a Pigment based inks Fig. 20b Dye based inks Printed colour inks onto a range of coated and uncoated papers. Thirty-seven papers were tested, including six coated papers - some of the more popular artists printmaking papers to which an ink receptive layer is added. Colour samples containing cyan, magenta, yellow, black and CMY 18 Over the course of a day the lux levels can change from 300 lux on an overcast winter morning, 400 lux for bright spring sunshine to 600 lux in the summer. 54

55 patches at 5% increments were printed using HP pigment inks and dyebased inks. Two sets of prints were made. The first set of printed test strips were mounted on paper sheets, were hung on a wall facing northwest and therefore the samples were not in direct sunlight, but to daylight conditions that simulated a gallery or home. 19 The second set was printed, allowed to dry for a day, wrapped in acid free tissue paper and black paper to prevent any exposure to light and humidity. These were used as a reference set for future comparison Photographing the samples At the time of the experiment, the colour patches were photographed using a SP-200xm metallurgical microscope that has a reflective light source, designed for non-transparent samples. Colour patches were photographed under x4 and x10 magnification using a Canon G3 digital camera. The microscope light source illuminated the surface of the sample from above. However, it was quickly found that the illumination was not constant, and was compounded by the Canon camera, resulting in a lack of colour consistency from sample to sample. The unsatisfactory results of the microphotography resulted in more of a reliance on E measurements rather than a visual comparison. 20 In order to address this problem, a discussion was held with experts as to a more suitable microscope and camera device for capturing micro images of ink on paper. Based on the design specification provided by microscopy expert Mary McCann and working with Nikon Company, a bespoke microscope was created comprising a Nikon SMZ800 stereo-microscope and a Nikon digital camera that was attached through a photo-tube beam-splitter. In order to view large print samples, the microscope body was mounted on a stand with a long arm extension. The samples were homogeneously lit by an 19 Based on guidelines for measuring light levels by M.McCormick of The Exhibition Alliance New York USA, a 35mm single reflex lens camera with in-built light meter is a fairly accurate measurement device. 20 This problem has taken some time to rectify and the purchase of a new microscope that was accurate and provided consistent colour recordings was only resolved through a meeting and discussion in 2008 with Mary McCann, an expert in microscopy who worked for the Polaroid Corporation (USA). 55

56 LED ring light, which ensured there were no shadows and the light source was consistent from one sample to the next. Therefore, a much higher degree of colour accuracy and consistency of colour could be achieved. Colour patches were photographed again in 2008 using the Nikon microscope. Cyan, magenta, yellow and black colour patches at 50% ink levels and a composite black patch at 100% were photographed. The camera package included the Nikon Elements software for viewing, capturing and for creating a library of samples Results from measuring and photographing the samples Delta E ( E) differences were recorded 21 between the original printed samples (reference data) and exposed samples (measured data). E averages were calculated to demonstrate the fading rates between CMYK, uncoated and coated, dye and pigment. Dye Cyan Magenta Yellow Black Coated Uncoated Pigment Cyan Magenta Yellow Black Coated Uncoated Table 6 Average E differences after 11 months between the original dye and pigment print samples and the exposed samples. As demonstrated in table 6, magenta, yellow and black dye on coated paper highlight the greatest E differences. All the dye samples indicate a much higher overall colour change, whereas the E difference measurements of pigment inks printed on coated and uncoated paper demonstrate less colour change, working from the E difference measurements and making a visual comparison of the samples. 21 A Gretag Macbeth spectrophotometer was used to measure the colour patches using an illumination D50 and at 2. The Centre used D50 or 5000 K for monitor calibration, viewing conditions and measuring (Wang & Parraman, 2003) 56

57 A Comparison of Samples Dye inks on coated papers Patches visible Intensity Composite Black and Black only Other cyan not visible on 5% and 10% ink patches; magenta patches up to 80% on some papers not visible. Colour intensity very reduced Composite blacks and greys appear olive green Colour mottling and non-uniform fading evident Dye inks on uncoated papers 5% patches still visible Colour intensity slightly reduced Composite blacks and greys of up to 75% appear greenish grey Table 7 A description of the changes to the colour patches over time. Pigmented inks on coated and uncoated papers 5% patches still visible Colour intensity only slightly reduced Slight reduction in intensity for composite black patches Papers that are coated are generally smoother and much brighter, and reflect colour more successfully than uncoated papers. For an uncoated paper, the fibres are still visible and the ink is displaced across the peaks and troughs of the fibres; the colour is reflected in different directions therefore making the image duller and flat. Uncoated papers are also more difficult to photograph, due to the uneven properties of the paper and fibre; the level of magnification and depth of field, some areas are sometimes out of focus. Using the Nikon microscope, the following micro photographs of dye and pigment samples printed onto coated and uncoated papers reveal two differences: the shape of the dots on the coated and uncoated papers, and the hue difference of the dye and pigment inks. As illustrated (figure 22a and 22b), the cyan pigment has a deeper and bluer hue than the cyan dye. The cyan and light cyan dots can be clearly discerned on the coated paper samples. The dots appear on the coated paper as more rounded and regular, whereas on the coated paper the dots 57

58 are less distinct and are absorbed into the structure of the paper fibre (figure 23a and 23b). In general, both samples on uncoated paper show no defined dot, there is less distinction between cyan and light cyan; and the pigment sample appears slightly paler. This is exaggerated in for example a Japanese Rayon and Tosa Shoji paper (figures 21a and 21b), which has a much looser weave, illustrating the network of fibres and how the ink is encapsulated within them. Fig. 21a Tosa Shoji 50 % black Fig. 21b Rayon Unryu 100 % black The samples that were photographed after 11 months of exposure, demonstrate a distinct colour change. After 11 months the 50% cyan dye has nearly completely faded (figure 22c) and just a few inkjet dots remain, whereas the light cyan dots on the fine art coated paper can be perceived to be slightly lighter. However comparing the magenta samples, the dye on fine art coated paper has completely faded (figure 24c). 58

59 Fig. 22a 50% cyan dye onto fine art coated. Fig. 22b 50% cyan pigment onto fine art coated. Fig. 22c 50% cyan dye onto fine art coated Fig. 22d 50% cyan pigment onto fine art after 11months. coated after 11 months. Fig. 23a 50% cyan dye onto fine art uncoated. Fig. 23b 50% cyan pigment onto fine art uncoated. 59

60 Fig. 23c 50% cyan dye on onto fine art uncoated after 11months. Fig. 23d 50% cyan pigment onto fine art uncoated after 11 months. Fig. 24a 50% magenta dye onto fine art coated. Fig. 24b 50% magenta pigment onto fine art coated. Fig. 24c 50% magenta dye on onto fine art coated after 11months. Fig. 24d 50% magenta pigment onto fine art coated after 11 months. 60

61 Fig. 25a 50% magenta dye onto fine art uncoated. Fig. 25b 50% magenta pigment onto fine art uncoated. Fig. 25c 50% magenta dye on onto fine art Uncoated after 11months. Fig. 25d 50% magenta pigment onto fine art uncoated after 11 months. The results suggest the magenta dye samples printed on coated papers appear to be particularly fugitive. From an inspection of the composite black colour patches (CMYK), as a result of the disappearance of magenta ink, results in faded colour samples that have a blue-ish hue to the samples regularly occur for example. Similarly for the uncoated papers the cyan is resistant to fading and therefore creates a similar greenish-blue result. For composite CMY (no black) samples, similar characteristics can be observed. The dots have receded in size and grown wider apart, light cyan and light magenta dots have faded and the magenta has near completely faded resulting in an olive brown colour (figures 26a and 26b). 61

62 Fig. 26a 100% black dye on fine art coated. Fig. 26b 100% black dye on fine art coated after 11 months Other indications that affect colour changes Colour measuring has also highlighted colour differences in which the paper sample has darkened. Subsequent measuring of the paper (Papanagiotou, 2000, p.73) has indicated that Japanese samples indicate a darkening or yellowing, suggesting that the fibre content can contribute to colour difference. Many of the accelerated lightfast tests yellowed the surface of coated papers, although this has not occurred during any real time tests. There was an assumption that ink on papers made by the same manufacturer would fade at the same rate. This has not been the case. Different manufacturing processes, such as the texture, sizing and construction of the paper can contribute to or inhibit the fading process. Another suggestion is the chemical effect of one colour on another. When printed separately a colour retains fidelity, however when combined with another is weakened and becomes more fugitive (Bugner and Artz, 2002, p ) Conclusion to the longevity of papers and inks This section investigated the permanence of printed inkjet samples. Whilst feeding could be perceived, a series of scientific methods were required to compare colour differences. The initial problem was to identify the types of methods that were appropriate to the artist. Traditional methods for comparing colour differences, as used by conservationists and scientists are 62

63 the Blue Wool Scale and E were most appropriate. Whilst these methods for measurements and comparison are useful, a visual comparison recording and inspecting the differences between pigment-based inks and dye-based inks on to a variety of coated and uncoated papers, provided a useful illustration of changes to the ink on paper over time. The printed samples were exposed to conditions that emulated exhibition and hanging conditions of fine art printed artwork. An investigation into the best possible way of photographically recording the samples resulted in the conclusion that a more reliable and uniform lit surface, and consistent methods to capture colour were required. Once the Nikon microscope and camera combination were put to use, more accurate colour images were obtained. Although E measurements were used, the photographed samples clearly demonstrated the colour and surface differences between coated and uncoated dye and pigment-based inks. 4.5 Conclusions to chapter 4 This chapter has highlighted and summarised a range of experiments that sought to address three problems relating to: inkjet printing on paper, appropriate colour management methods for printing, and recording of the fading of ink on paper. What was particularly apparent during this time was an emerging scientific over-reliance on the colorimetry of colour rather than colour and appearance. My interests were largely based on colour appearance: colours that were pleasing to the eye but not necessarily accurate; the subjective relationship of one colour to another, and their psychological impact. As the hardware and software technology has significantly developed, there has been an improvement in reducing the margin of error between the transfer of data between devices. However, in turn this has removed the hands-on opportunity to engage in the serendipitous happenings and decision making required when colour mixing paints and pigments. The idea to create an alternative colour palette for inkjet machines was based on the screenprint tests undertaken at the beginning of the chapter. This emerging idea 63

64 demonstrated that whilst CMYK inks presented the best method for the reproduction of RGB colours, the notion of tools for creative colour mixing has been overlooked. What can be considered as important is the difference between additive and subtractive colours, between process colours and artists colours, and the differences in colour models to plot each colour necessary for each. In order to decide on a suitable alternative colour system, the artist could benefit from: The addition of colours to the existing CMYK colour range; or By creating a separate colour set that corresponds to artists colours; as one might mix from a series of pots or tubes in a studio. In order to form a basis for selecting a colour palette, one might consider the differences in colour gamut. For example, when added in equal measures, CMYK inks are designed to provide as close as possible to the primary spectral colours of red, green and blue, therefore reproducing the colours on a computer screen. By removing the CMYK colours from the colour set, and add other non-process colours, we may reduce some areas of the gamut, but expand in a different part of the colour space. When printing with traditional artists pigments, for example in intaglio, screenprint and litho, these result in a slightly more reduced colour gamut but present other considerations such as extenders for translucent colours, the use of white to increase opacity, and the opportunity to over-layer colour, which cannot be easily undertaken in the inkjet process. Alternative colours may assist in improving inkjet printing by firstly demonstrating how colours are reproduced from what is seen on computer screen: how paintings, for example, can be more faithfully reproduced in books or as posters (print on demand); and secondly how an artist makes an inkjet print. 64

65 5. Other experiments relating to Chapter 4 of the thesis: The development and visualisation of bespoke colour sets for inkjet printing 5.1 Early experiments into the creation of colour charts Early investigations involved exploratory experimentations using a range of artists drawing ink colours to ascertain whether traditional drawing inks could be used in print cartridges and how these colours printed and combined on paper. The objective was to ascertain whether the thermal printheads could be filled with any type of ink and to develop methods for refining the process of filling print-heads. Draw-downs of inks using a K-Bar were undertaken and a visual assessment was made. A Windsor and Newton range was tried, however they are highly unsuitable as these inks contained shellac, which fused to the inside of the print-head and blocked the nozzles. The Dr Johnsons inks do not contain shellac and the colours are very intense. However, the inks do contain tiny particles, which could potentially block the nozzles. It was therefore important to rest the ink in the bottle, to allow the sediment to fall to the bottom, and not to disturb the liquid when drawing the ink out of the bottle. The use of the Dr Johnson inks was relatively successful. The colours were highly saturated and provided good ink coverage (figure 27). However due to the accidental inclusion of particles in the ink, the printhead nozzles often became blocked. Following from the Dr Johnsons inks, it was decided to use HP Vivera inks (see main thesis 4.2.1), as these already had the guaranteed purity for inkjet printing and could be mixed without the problem of clogging. It was decided to concentrate on mixing new colours using the existing HP inks and to test whether more interesting colours could be created when printed. As described in chapter 4.2 of the main thesis, a method was devised for making draw-downs of inks, measuring and sorting hues that could complement each other within an inkjet palette. In order to identify the correct transition of light to dark and placement of hue colours in the print 65

66 channels, a lengthy series of print tests were undertaken. The following charts (figure 27) demonstrate how, by combining some of the process colours with newly mixed colours and rearranging their order, some visually interesting colour sets began to be achieved. The most successful arrangements demonstrated a deep richness of colour and smooth transition between hues and tone. Dr J colour chart July 5th 2005 (15) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Mixed HP Inks hues Orange Brown Cherry Red Violet Persimmon True Blue Dr J colour chart July 6th 2005 (2) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Mixed HP Inks hues Orange Brown Scarlet Violet Persimmon True Blue Dr J colour chart July 7th 2005 (4) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Mixed HP Inks hues Orange Brown Magenta Violet Persimmon Scarlet Fig. 27 Some examples of the charts produced using the Dr Johnsons inks. A scan of the printed chart can be seen in the right box, the names of the ink colours are listed in the middle column and their placement according to the channels in the printer are listed in the left column. 66

67 5.2 Printer hardware and settings Fig. 28 HP130nr, 24 printer. HP Thermal Inkjet Print technology, with a resolution of 2400 x 1200 dpi (Resolution technology HP Color Layering technology, PhotoREt IV). There are 6 colour cartridges (1 each black, cyan, magenta, yellow, light cyan, light magenta), which uses HP Vivera fade-resistant dye-based inks. There are 304 printhead nozzles per head, with an ink droplet size of 4 picolitres. Hardware PowerBook G4 MAC OS X Version Imaging software Photoshop 7 Colour settings/working space Adobe RGB (1998) Source Space Untagged RGB Print settings Same as Source HP Printer Driver software HP Inkjet Driver version Papertype/profile HP Premium Plus Photo Satin and Hahnemühle Photo Rag matte Quality Best Colour Application Driven Color Source Roll Computer to printer connection USB Table 8. In order to maintain parity for all prints, the same working space and printer settings were used. 5.3 Colour charts and three-dimensional plots A method was devised to systematically record all the variations and the placement of the mixed colours in the channels as a new colour chart was created. The printheads were re-used several times and moved between the channels. When the printhead was moved to a different channel the printer chip on the printhead had to be removed and replaced with the correct printer chip. The column on the left listed the six HP printer channels, the second middle column listed the newly mixed colours, an empty box in this middle column indicated that a process colour was used. A scanned version 67

68 of the printed chart was included as a reference guide, which is shown in the right box. In the following figures the dominating colours are orange, bright red and deep red. However the transition from red to the vertical strips of violet and brown and from light red to dark red are not smooth with an over dominance of red that has a small colour difference. Possibly the orange could be placed in the light magenta channel, the red in the magenta channel and a light violet in the light cyan channel. Figures 30a and b demonstrate how, by swapping the cyan and the black cartridges, different colour effects can be achieved HP Red colour chart HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Mixed HP Inks hues Blue Violet Red Orange Fig. 29 An example of a printed colour chart using HP mixed inks Vivera inks colour chart July 7th 2005 (5) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Fig. 30a Mixed HP Inks hues Blue Violet Red Orange Vivera inks colour chart July 7th 2005 (6) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan HP Channels Violet Blue Red Orange Fig. 30b By swapping the cyan and the black cartridges different colour effects can be achieved 68

69 The following three figures 31a - c show how, by moving the premixed colours to different channels, very different colour mixtures can be obtained. The last table demonstrates a much better transition of colour. Vivera inks colour chart July 5th 2005 (12) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Fig. 31a Mixed HP Inks hues Violet Blue Orange Vivera inks colour chart July 5th 2005 (13) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Fig. 31b Mixed HP Inks hues Blue Violet Orange Vivera inks colour chart July 5th 2005 (1) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Mixed HP Inks hues Violet Orange Light blue Fig. 31c By moving the premixed colours to different channels different colour mixtures can be obtained. The figures 32a and 32b demonstrate a more systematic approach to ink mixing. Here light and dark pairs are produced to create a smoother tonal transition from light to dark, in order to remove less of the obvious jumps that appear in the earlier test prints. The blue ink in the black cartridge is 69

70 replaced for a violet in the subsequent chart (figure 32b), thus creating a brown, instead of a dark green. By heightening the light magenta cartridge to a light red instead of a light orange, there is more of a shift from green midcolours to brown mid-colours (figure 34). In figure 35, here a black and yellow remains. A jump appears from light magenta to magenta, which has a purple in the magenta channel. In order to obtain the smooth transition, violet and light violet, rose and light magenta pairs are used, as seen in figure 36. Vivera inks colour chart July 28th 2005 (8) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Fig. 32a Mixed HP Inks hues Dark yellow 2Y:1C Blue 1LM:2C Rose 2M:1Y Green 2Y:1C Light Green 2Y:1LC Vivera inks colour chart July 27th 2005 (7) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Fig. 32b Mixed HP Inks hues Dark yellow 2Y:1C Violet 1M:1C Rose 2M:1Y Green 2Y:1C Light Green 2Y:1LC Vivera inks colour chart July 27th 2005 (11) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Fig. 33 Mixed HP Inks hues Purple Red Blue Orange Light Blue 70

71 Vivera inks colour chart July 27th 2005 (17) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Fig. 34 Mixed HP Inks hues Purple Red Blue Light red Light Blue Vivera inks colour chart July 27th 2005 (16) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Fig. 35 Mixed HP Inks hues Purple Blue Red Light Blue Vivera inks colour chart July 27th 2005 (18) HP Channels Yellow Black Magenta Cyan Light magenta Light cyan Mixed HP Inks hues Purple Rose Violet Light Violet Fig. 36 Examples of a more systematic approach to ink mixing and refinement of the process The following section demonstrates the results of the refinement process. Each colour chart includes the table of colours with the printed and scanned reference, along with the measured ICC profile and its plot in a threedimensional space as viewed from the lightest viewpoint and the darkest viewpoint (see chapter in the main thesis). The third plot describes the measured CMY primaries and compares these points to the original CMY cube (see chapter in the main thesis). 71

72 RGB-CMYK colour chart: 1. rose-violet-dark-yellow HP Channels Custom colours Yellow Dark Yellow Black Purple Magenta Rose Cyan Violet Light magenta Light Magenta Light cyan Light Violet 72

73 RGB-CMYK colour chart: 1. rose-violet-dark-yellow HP Channels Custom colours Yellow Dark Yellow Black Purple Magenta Rose Cyan Violet Light magenta Light Magenta Light cyan Light Violet 73

74 Chart 1. rose-violet-dark-yellow Multi view point and from above Y K M C K M C Y 74

75 Chart 1. rose-violet-dark-yellow View from the front and the left K Y M C K Y C M 75

76 RGB-CMYK colour chart: 2. rose-violet HP Channels Custom colours Yellow Yellow Black Purple Magenta Rose Cyan Violet Light magenta Light Magenta Light cyan Light Violet 76

77 RGB-CMYK colour chart: 2. rose-violet HP Channels Custom colours Yellow Yellow Black Purple Magenta Rose Cyan Violet Light magenta Light Magenta Light cyan Light Violet 77

78 Chart 2. rose-violet - Multi view point and from above K Y M C K M C Y 78

79 Chart 2. rose-violet View from the front and the left K Y M C K Y C M 79

80 RGB-CMYK colour chart: 3. red-violet HP Channels Custom colours Yellow Yellow Black Purple Magenta Red Cyan Violet Light magenta Orange Light cyan Light Violet 80

81 RGB-CMYK colour chart: 3. red-violet HP Channels Custom colours Yellow Yellow Black Purple Magenta Red Cyan Violet Light magenta Orange Light cyan Light Violet 81

82 Chart 3. red-violet - Multi view point and from above K Y M C K M C Y 82

83 Chart 3. red-violet View from the front and the left K Y M C K Y C M 83

84 RGB-CMYK colour chart: 4. blue-green-black HP Channels Custom colours Yellow Yellow Black Black Magenta Blue Cyan Green Light magenta Light Blue Light cyan Light Green 84

85 RGB-CMYK colour chart: 4. blue-green-black HP Channels Custom colours Yellow Yellow Black Black Magenta Blue Cyan Green Light magenta Light Blue Light cyan Light Green 85

86 Chart 4. blue-green-black - Multi view point and from above K Y C M M K Y C 86

87 Chart 4. blue-green-black View from the front and the left K Y M C K Y C M 87

88 RGB-CMYK colour chart: 5. red-green-black HP Channels Custom colours Yellow Yellow Black Black Magenta Red Cyan Green Light magenta Orange Light cyan Light Green 88

89 RGB-CMYK colour chart: 5. red-green-black HP Channels Custom colours Yellow Yellow Black Black Magenta Red Cyan Green Light magenta Orange Light cyan Light Green 89

90 RGB-CMYK colour chart: 6. red-green HP Channels Custom colours Yellow Yellow Black Purple Magenta Red Cyan Green Light magenta Orange Light cyan Light Green 90

91 RGB-CMYK colour chart: 6. red-green HP Channels Custom colours Yellow Yellow Black Purple Magenta Red Cyan Green Light magenta Orange Light cyan Light Green 91

92 Chart 5. red-green-black and Chart 6. red-green - Multi view point and from Above K Y C M M K Y C 92

93 Chart 5. red-green-black and Chart 6. red-green View from the front and the left K Y M C K Y C M 93

94 RGB-CMYK colour chart: 7. rose-green HP Channels Custom colours Yellow Yellow Black Purple Magenta Rose Cyan Green Light magenta Light Magenta Light cyan Light Green 94

95 RGB-CMYK colour chart: 7. rose-green HP Channels Custom colours Yellow Yellow Black Purple Magenta Rose Cyan Green Light magenta Light Magenta Light cyan Light Green 95

96 Chart 7. rose-green - Multi view point and from above K Y C M K M Y C 96

97 Chart 7. rose-green View from the front and the left K Y M C K Y M C 97

98 RGB-CMYK colour chart: 8. red-blue HP Channels Custom colours Yellow Yellow Black Purple Magenta Red Cyan Blue Light magenta Orange Light cyan Light Blue 98

99 RGB- CMYK colour chart: 8. red-blue HP Channels Custom colours Yellow Yellow Black Purple Magenta Red Cyan Blue Light magenta Orange Light cyan Light Blue 99

100 Chart 8. red-blue - Multi view point and from above K Y M C K M C Y 100

101 Chart 8. red-blue View from the front and the left K Y M C K Y M C 101

102 RGB-CMYK colour chart: 9. rose-blue HP Channels Custom colours Yellow Yellow Black Purple Magenta Rose Cyan Blue Light magenta Light Magenta Light cyan Light Blue 102

103 RGB-CMYK colour chart: 9. rose-blue HP Channels Custom colours Yellow Yellow Black Purple Magenta Rose Cyan Blue Light magenta Light Magenta Light cyan Light Blue 103

104 Chart 9. rose-blue - Multi view point and from above K Y M C K M C Y 104

105 Chart 9. rose-blue View from the front and the left Y K M C K Y M C 105

106 RGB-CMYK colour chart: 10. green-blue HP Channels Custom colours Yellow Yellow Black Purple Magenta Green Cyan Blue Light magenta Light Green Light cyan Light Blue 106

107 RGB-CMYK colour chart: 10. green-blue HP Channels Custom colours Yellow Yellow Black Purple Magenta Green Cyan Blue Light magenta Light Green Light cyan Light Blue 107

108 Chart 10. green-blue - Multi view point and from above K C Y M M K Y C 108

109 Chart 10. green-blue View from the front and the left K Y M C K Y C M 109

110 RGB-CMYK colour chart: Ruth 3. red-green HP Channels Custom colours Yellow Yellow Black Purple Magenta Red Cyan Green Light magenta Light Magenta Light cyan Cyan a) Measurement of colour chart, and three-dimensional plots (side elevation) 110

111 b) Measurement of colour chart, and three-dimensional plots (side elevation) 111

112 c) Measurement of colour chart, and three-dimensional plots (view from above) 112

113 d) Measurement of colour chart, and three-dimensional plots (view from below) 113

114 RUTH - Multi view point and from above K Y C M K M Y C 114

115 RUTH View from the front and the left K Y M C K Y M C 115

116 5.3.1 How and where to install the profiles On a Mac computer, this is undertaken by firstly going to: Hard drive> Users (Users own file space) > Library > Colorsync > Profiles; and copying the files to this folder (figure 37). Fig. 37 List of profiles in the ColorSync folder For other operating systems the profiles should be copied to the following folders (table 9). Then, restart the image application software, eg. Adobe Photoshop Operation System Macintosh OS 9 Macintosh OS X Windows 9x/Me WindowsNT Windows 2000/XP/2003 ICC profiles should be copied into: System folder\colorsync profiles Users\library\colorsync\profiles Windows\system\color WinNT\system32\color WinNT\system32\spool\drivers\color Table 9 System folders on other operating systems Using the profiles Open Photoshop or an image application software. Open an image file. The TC9.18 RGB.tiff target might be the best to give an indicator of the colour changes. Then, in order to change the colour, a profile needs to be assigned. If placed in the right folder, these profiles can be found by going to Image > Assign Profile > check the Profile button and access the pull down menu. (figures 38a, b and c). 116

117 Fig. 38a Fig. 38b Fig. 38c The procedure for assigning a profile to soft preview the colours. If the profiles do not appear, try quitting Photoshop and then reopen. If this is unsuccessful, ensure the profiles are copied correctly to the right library folder in Users section. 117

118 5.4 A description of the colour test charts A range of colour test charts was designed for specific colour printing requirements, which was refined as the research evolved. These templates were designed for visualising colour relationships (on screen and on paper), to demonstrate alternative methods for colour mixing, colour printing, and over-layering of colours and pattern. Chapter 5 provides more information on templates 1 and 2. Each template is briefly described here: Template 1: Blended squares The first experiment investigated the blending of the four primary channels: cyan, magenta, yellow and black in order to ascertain how the inkjet printer drives the primary colours in one pass. The first template therefore included the four primary colours. Each corner represented a 100% value of each of the main channels and each colour was blended to white towards the centre. A chart was generated with layers, using the multiply function to enable all the layers to visibly mix. Any bespoke inks that are inserted into the printheads will be printed with the corresponding primary inks. Fig. 39 Three variations of CMYK blends. 118

119 5.4.2 Template 2: Blended six-pointed stars In order to incorporate the light magenta and light cyan channels a six pointed star was devised. With the benefit of a previous test into pinpointing how the primary hue and lighter version overlapped, each corner represents 100% values of all six channels in the printer: cyan, magenta, yellow, black, light cyan, light magenta, black. This was used in conjunction with the RGB printer to ascertain the relationship of the custom colours and CMYK colours. To obtain a twelve pointed star, the hexagonal star was rotated on its axis by 90. Fig. 40 The creation of the blended stars in Photoshop Template 3: To assess the appearance of printed colours through the overprinting of solid colours With the help of a third template devised for this research, the increasing density levels of a colour were demonstrated as each custom primary colour is overprinted. Three of the squares tested an overprint of the same colour, and the fourth (bottom left) assessed the effect when the light magenta and magenta channels were combined (figure 41). Using this template a printhead filled with any non-standard colour can be tested. The following section describes the construction and method for printing the chart in more detail. 119

120 Fig. 41 The layers option is used to print the separate layers of colour. An ICC profile rosegreen is used to soft preview the final printed colours, for example, the cyan channel is viewed here as green. Individual layers are selected and deselected according which layer is printed Method for constructing template 3 Working in Photoshop, an image was constructed using the layers option. A test sheet was designed with 4 squares of 1400 x 1400 pixels decreasing by 280 pixels, as combinations of 100% magenta, light magenta, yellow, and cyan. Each layer represented one pass through the printer; in this instance the paper passed through the printer five times. As illustrated in figure 41, the thumbnail images in the layers palette show the position of the squares. Note also, the square in the bottom left hand corner is a composite of 100% light magenta, magenta and yellow Method for printing the template 3 Using the HP Designjet 130nr printer with a set of mixed ink colours (rose, light magenta, dark yellow, green, light green, and purple), individual layers were overprinted and passed through the printer five times (figure 42). Layers were selected and deselected according to which layer was printed. Using an ICC profile rose-green the mixed colours rose, dark yellow and green were assigned to the RGB magenta, yellow and cyan channels. The 120

121 print space remained as same as source with paper-type set to HP Premium Heavyweight Coated, with other settings as: best colour rendition and application managed colour. a) b) c) d) Fig. 42 An example of overprinting of coloured squares using pink, deep yellow and green primary hue colours. The three squares a, b and d are each overprinted using only red for a; dark yellow for b; and green for d. (Square c) is a composite of dark yellow, light magenta and red). By overprinting with the same colour a denser, richer colour can be obtained. These denser, richer colours, so far, cannot be obtained by any other means. Following a discussion as to whether there was any perceivable difference to the printed colour if the sequence of colours was changed, modifications were made to the order of the overprinting of the colour layers, and three different versions were printed (the first layer was printed last or vice versa). Printed version1: The largest squares were printed first; the smaller squares were overprinted; the last square being the smallest. The black layer (smallest square) was not printed in this print. Printed version 2: Starting with the black layer, the smallest squares were printed first, then overprinting to the largest square Printed version 3: Largest squares first, then over-layering to smallest squares. 121

122 Measuring the spectral reflectance density of colour After measuring and comparing the three printed versions, there is very little difference, with a range from E difference in the lighter and mid colour areas, with a maximum of a E difference of 4.4 in the darkest squares. This may be due to a slight amount of scuffing to the surface of the paper by the printer heads Template 4: To assess the appearance of printed colours through printing patterns on top of solid colours A fourth chart is devised specifically for this research to plot the relationship between each of the CMYK printer channels and the RGB colour space. A nine-by-nine square chart was generated, and at each corner 100% values of all custom primary hues in an RGB colour space were represented (figure 43). The chart was used in particular to overprint a combination of coloured squares using custom colours, inkjet patterns and screenprinted patterns. This chart can also be used for overprinting using patterns (figure 44) of colours or solid colours. Comparisons are made by overprinting patterns onto printed colour squares using different colour sets, such as printing a green and rose ink set on top of a blue and red ink set. This method is described in more detail in chapter 6 in the main thesis. 122

123 Fig. 43 A graded square with the four primary colours at each corner: cyan, magenta, yellow and black Fig. 44 Example of the patterns that are printed over solid colours 123

124 5.5 Visualising and comparing the mixed inks: Results from chapter of the main thesis Chart 1 L A B C M Y C M Y C+M C+Y Y+M C+Y+M Chart 2 L A B C M Y C M Y C+M C+Y Y+M C+Y+M Chart 3 L A B C M Y C M Y C+M C+Y Y+M C+Y+M

125 Chart 4 L A B C M Y C M Y C+M C+Y Y+M C+Y+M Chart L A B C M Y 5+6 C M Y C+M C+Y Y+M C+Y+M Chart 7 L A B C M Y C M Y C+M C+Y Y+M C+Y+M

126 Chart 8 L A B C M Y C M Y C+M C+Y Y+M C+Y+M Chart 9 L A B C M Y C M Y C+M C+Y Y+M C+Y+M Chart 10 L A B C M Y C M Y C+M C+Y Y+M C+Y+M

127 Ruth L A B C M Y C M Y C+M C+Y Y+M C+Y+M

128 6. Supporting documentation for Chapter 5: An artist s introduction to inkjet: A case study and a generic working model The following section provides supporting documentation for the generated prints made with Ruth Piper and lists the methods, colours and order of printing. Fig. 45 Composite layers in Photoshop of version 1 of Hub Cap: FRD Version 1 of print 1 Hub Cap: FRD 857, November 2006 The order of printing: Print pale yellow outer ring (layer 2) and bright yellow disk (layer 4); Print blue ring (layer 42) and spokes (layer 43); Print blue ring (layer 42) and spokes (layer 43); Print red ring (layer 3) and red dots (layer 35); Print purple disk (layer 4 copy); Print purple disk (layer 4 copy). 128

129 6.1.1 Version 2 of print 1 Hub Cap: FRD 857 December 2006 On reflection, Ruth decided to change the central rings of the print. Keeping the first three actions the same, an extra central cog was added and printed in a blue-grey, the orange ring was replaced by a pale olive green, and an extra ring of pale blue grey was added, to sit on top of a pale olive green. The order of printing: Print pale yellow outer ring (layer 2) and bright yellow disk (layer 4); Print blue ring (layer 42) and spokes (layer 43); Print blue ring (layer 42) and spokes (layer 43); Change the orange ring to a pale olive green, colour Q8, and print with the orange dots (layer 35); Print deep purple twice (layer 4); Print background (layer 45) with small blue grey cog using colour X9 (layer 48); Print blue grey thin circle X9 (layer 46) and grey cog again Print 2 version 1, Caprice CHV954, January 2007 In order to obtain the turquoise blue segments, the image was first printed through the CMYK printer. Both the background and the turquoise were printed. We discussed an approach to incorporating the incidentals, such as the creases in the paper, the buckle of the drafting paper on which the cut sections were applied and the shadows created by the edges of the cut paper (figure 46a). Piper was also interested in selecting areas using the wand so that edges were less clearly defined. The appearance of this print is less hard-edge than the first. Again, in order to obtain the brightest colours, holes were punched through to the white, with layers or group of layers printed individually. The order of printing: Print turquoise sections and background using CMYK colour J23, (layer 13) in srgb printer managed colour; 129

130 Return to custom colour printer using application managed colours, print black dots (layer 1) (layer 14) and (layer 7); Print (layer 15) and (layer 16); Print (layer 14), (layer 12 copy), (layer 7). Fig. 46a Version 1 Caprice CHV954 Fig. 46b Version 2 Caprice CHV Print 2 version 2, Caprice CHV954, January 2007 Small changes are made to the second version (figure 46b): extra dark yellow lines are removed from the central yellow disk; the black dots are replaced by red dots; to obtain red dots that are as red as possible the hue and saturation tool is used to the maximum amounts. The order of printing: Print turquoise sections and background using CMYK colour J23, (layer 13) in srgb printer managed colour; Return to custom colour printer using application managed colours, print red dots (layer 1) (layer 14) and (layer 7); Print (layer 15) and (layer 16); Print (layer 14), (layer 12 copy), (layer 7). 130