Newsshade 2003
02 Foreword 2004 Ifra, Darmstadt Foreword In 1995, Ifra published a Special Report 1.11 in which a shade for standard newsprint called Newsshade 95 was defined. This shade was a modification and an update of the earlier shade, which had been promoted by the Scannews Technical Committee and adopted by Ifra with the designation Newsshade 88. Now, a new research project has just been completed which sets out to update the values associated with Newsshade 95 by taking into account technical advances in both measurement methods and production processes. The project also included a survey of improved newsprints, since the improved grades account for an increasingly large share of the paper currently used by newspaper printing plants. Many supplements, and also the front pages of some newspapers, are today printed on this paper grade. Measurements have been carried out, the results have been analysed and the data have been discussed. These data refer not only to standard newsprint, but also to a large number of improved newsprint and Value Added Coldset samples. The results show that a wide range of shades and a wide range of brightness values can be found among the improved newsprint samples. It is not however possible to group the samples into any significant clusters, and the working group decided that the values for the improved newsprint should be further discussed before any recommendation is made concerning a standard Improved Newsshade. Newsshade 2003 is defined only for standard newsprint. This standard is based on a description of the newsprint shade in the CIELAB colour space and, provided that the illuminant is carefully selected and the measurement conditions are clearly defined, it provides a definition of a unique and reproducible shade. The approach taken in this project was not to arbitrarily define a shade and impose it on the paper mills. That would be neither realistic nor loyal towards those partners who have long followed the recommendations embodied in previous newsprint standards. The first step was to agree upon a method of measurement that is in accordance with the latest technological developments and that satisfies the preconditions for developing such a standard and that is, if possible, available at the paper mills. Samples from a wide range of different papers available on the market were then collected and their optical properties were measured. The statistical analysis of the results obtained allowed several statistically representative average shades to be identified among this collection of samples. This provided the basis on which the shades described in this report were established, and ensured that the standard shade thus defined represents products actually available on the market. Under the guidance of the Ifra Material Management Committee, a working group was formed to lead the project. The members of the working group were, in alphabetical order, as follows: Dr. J. Anthony Bristow, Bristow Consulting AB, Tullinge, S Dan Buskhe, Holmen Paper, Norrköpping, S Hans Ekblad, Lorentzen & Wettre, Kista, S Laurent Favier, SPPP, Paris, F Roland Geipel, Datacolor GmbH, Marl, D Prof. Dr. Lothar Göttsching TUD, Darmstadt, D Heikki Helevuo, Stora Enso Newsprint Summa Mill, Hamina, Fin Anders Hillvall, SCA Forest Products, Sundsvall, S Fried Hoenderboom, PCM Grafisch Bedrijf, Amsterdam, NL Dr. Erwin Krauthauf, UPM-Kymmene Papier GmbH & Co. KG, Schongau, D Sebastian Loewenberg, UPM-Kymmene Papier GmbH & Co. KG, Schwedt, D Anton Obermeier, Lorentzen & Wettre, Munich, D Timo Palo, UPM-Kymmene, Helsinki, Fin Dr. Hermann Praast, TUD, Darmstadt, D Kaisu Soudant, UPM-Kymmene, Kajaani, Fin Peter Terins, Holmen Paper AB Bravikens Pappersbruk, Norrköping, S Bjørn Thomassen, Norske Skogindustrier ASA, Skien, N The measurements were carried out at IfP (Institut für Papierfabrikation), Darmstadt, under the guidance of Prof. Dr. Lothar Göttsching and Dr. Hermann Praast. We take this opportunity to thank these persons for their valuable contribution to the project. Ifra was represented by Manfred Werfel, Director of Research, Consulting and Training, and by Eric Languillat, Frédéric Fabre and Beatrix Beckmann, Ifra research engineers. This report is a summary of the project work, and it is intended to provide all the information necessary to the various parties likely to make use of these results, while avoiding details of those steps in the project that are likely to be a burden on the reader without bringing any genuine added value to the report. Of course, all the information is at the disposal of those who wish to consult it. Anyone wishing to do so is requested to contact Ifra. March 2004
2004 Ifra, Darmstadt Table of contents and Imprint 03 Table of contents Foreword...03 1 Introduction...04 2 Terminology...04 3 Choice of measurement conditions...05 3.1 Type of instrument...05 3.2 Type of sample...05 3.3 Illuminant/observer conditions...05 4 Measurement procedure...06 4.1 Instrument...06 4.2 Measurements...06 4.3 Calculated parameters...06 5 Results...07 5.1 Standard newsprint...07 5.1.1 Newsshade 2003...07 5.1.2 Marketshade 2001...09 5.2 Other measurement results...09 5.2.1 Effect of fluorescence...09 5.2.2 CIE whiteness (C/2 )...11 6 Colour designations under D50 conditions...12 6.1 Colour designations under CIE illuminant D50 conditions...12 6.2 Colour designations under D50 conditions over a black backing...12 6.3 The requirements of ISO 12647-3...13 7 Conclusion...14 8 References...15 Imprint Ifra Special Reports, research reports, technical study reports and documents for the standardisation of newspaper production techniques. Published by: Ifra, Washingtonplatz, 64287 Darmstadt, Germany; www.ifra.com; Tel. +49.6151. 733-6; Fax +49.6151.733-800. Chief Executive Officer: Reiner Mittelbach. Director of Research and Consulting: Manfred Werfel. Research Manager: Harald Löffler. Republishing - also of excerpts - only with express permission of Ifra and acknowledgement of origin. Price: Ifra Special Reports are sold at the price of 130 EUR per copy. For Ifra members, the price is covered by the membership fee that entitles them to an allotted number of copies. Ifra members may order additional copies at 13 EUR per copy. * plus 7% in Germany and for companies and persons in the European Union that do not have a VAT number.
04 1 Introduction / 2 Terminology 2004 Ifra, Darmstadt 1 Introduction In 1995, Ifra presented specifications that could be used by newspaper printers when ordering the standard newsprint shade designated Newsshade 95 from paper mills. Now it has been felt that there is a need to update these specifications and also to include other paper grades for cold-set printing, such as the so-called improved newsprint. It has, however, been realised from the beginning that this ambition may be difficult to achieve. The acceptable visual tolerances are so narrow that, as indicated in the 1995 report relating to Newsshade 95, differences in calibration standards and differences in the level of UV-adjustment in different measuring instruments can create major problems. In addition, there are differences in the measuring conditions traditionally used within the papermaking industry, on the one hand, and the printing industry, on the other. These questions are discussed in more detail later in this report. Paper grade Grammages [g/m 2 ] Number of samples Newsprint 40, 42, 45, 48.8 97 Improved newsprint 52, 55 59 VAC (Value Added Coldset, all available see Ifra Special Report 1.20) grammages 2 Table 1: Paper grades studied It was decided to begin by studying the shades of newsprint currently available around the world and to compare them with the existing standard. This would provide data to enable a Marketshade 2001" to be identified and the magnitude of the variation of standard newsprints around this centroid value to be assessed. It would also indicate to what extent it is necessary to update the coordinates of the Newsshade standard. At the same time, other grades of newsprint would be collected and measured, and a second stage in the project would be to statistically analyse the distribution of these grades with regard to shade, in order to see whether it was possible to identify an improved newsprint most commonly used by our industry. A major result was that it was not possible to define a standard improved newsprint within the project. Further analyses and perhaps further developments will be necessary before this goal can be achieved. The working group of experts met on two occasions in Darmstadt; the first time to decide which measurements should be made and how they should be made, and the second time to consider the data obtained by these measurements and to decide what information could be derived, what conclusions could be drawn and what proposals could be made on the basis of the analysis. A total of 97 standard newsprints, 59 improved newsprints and 2 VAC (Value Added Coldset) samples were received from different paper mills in Europe, North America and Asia. The study was limited to the grammages listed in Table 1 for the different grades. 2 Terminology There is reason to believe that some readers of the previous Newsshade report may have been confused by the duality of the project. The previous study was indeed based on a collection of measurements intended both to show the shade of the newsprints available on the market and to define the standard shade of newsprint most commonly used. As a result, the term Newsshade 95 has been used as both (a) a general statistical description with wide confidence limits of the grades of newsprint in current use and (b) a specification with narrow tolerances of a specific grade of newsprint. The present report also has a dual purpose. It seeks to provide a statistical picture based on an analysis of samples received from all around the world and, at the same time, it seeks to present realistic values to define the new standard. In order to avoid further confusion, the term Marketshade 2001 is here introduced to refer to the statistical material and the term Newsshade 2003 to designate the specific standard shade.
2004 Ifra, Darmstadt 3 Choice of measurement conditions 05 3 Choice of measurement conditions A number of questions regarding the type of instrument, the type of sample, the measurement conditions and the type of calculation to be adopted had to be considered by the working group before it was possible to make any measurements. 3.1 Type of instrument The paper industry makes measurements according to rigorous standards defining measurement and calibration conditions using a d/0 geometry (ISO 2469 etc) with instructions for adjusting the UV-content of the illumination to a specified level. In contrast, the printing industry commonly uses a 0 /45 or 45 /0 geometry with less rigid calibration instructions and no UV-control (ISO 13655). After some discussion, it was decided in the present study to begin with the papermaker and to take steps to ensure that all the reference measurements were made under the best possible conditions in an instrument conforming to ISO 2469, the standard to which all newsprint mills relate their measurements. Further, it was agreed that the instrument should be calibrated in accordance with the provisions of ISO 2469 using an IR3 1 reference standard provided by the ISO/TC6 2 -authorized laboratory at STFI, Stockholm. Similarly, the UV-adjustment should be made with a fluorescent reference standard from the same laboratory, in accordance with the requirements of ISO 2470 and ISO 11475. In the subsequent analysis of the data, attempts would be made to convert the data to a form more recognisable within the printing industry. 3.2 Type of sample The shade of newsprint as determined by the papermaker is grammage-independent and is defined in terms of measurements made on an opaque pad of paper. This is a relevant measure for the printer ordering a supply of newsprint. For practical reasons, however, the printer usually makes measurements not on an opaque pad but on a single sheet of paper over a black background. This leads to a value which is grammage-dependent and opacity-dependent. It was decided that the initial measurements should be made according to ISO 2469 on an opaque pad of paper, but that informative single-sheet data should, if possible, be included in the report. 3.3 Illuminant/observer conditions Colorimetric data are always related to a stated illuminant/observer combination, such as C/2, D65/10 or D50/2. Here, again, there is a difference between the papermaker who traditionally reports data related to C/2 conditions and the newspaper printer who prefers the D50/2 (= ISO 12647, process control standard) combination. Some commercial printers also use D65/10 conditions. Here, however, a distinction must be made between the choice of measurement conditions and the choice of calculation conditions. In modern spectrophotometric instruments, the only measurements made are so-called spectral measurements, i.e. reflectance factor measurements at different wavelengths over the visual range from 400 to 700 nm. The choice of observer conditions (2 or 10 ) can always be included in the calculations after the measurements have been done. There is, however, a problem with the choice of illuminant. In many cases, this choice can also be made after the measurements have been made but, if the paper exhibits fluorescence, the UV-content of the illumination must be adjusted to match the desired illuminant before the measurements are made. Such an adjustment was necessary in this present study, since modern newsprints particularly the improved grades often exhibit fluorescence. The choice was limited, because standard procedures have been developed in relation to the d/0 instruments for adjustment to UV(C) conditions and UV(D65) conditions, but no such procedures for adjustment to UV(D50) conditions are yet available. As already indicated, the primary decision was to measure in accordance with the paper industry s standards with an adjustment to match the C-illuminant. It was, however, agreed that some complementary measurements would be made with an adjustment to UV(D65) conditions, and also that measurements would be made in the UVex(420)-mode, where a 420 nm cut-off filter is introduced into the light beams to eliminate all the fluorescence phenomenon. 1 ISO reference standard of level 3 2 Technical Committee Paper, boards and pulps
06 4 Measurement procedure 2004 Ifra, Darmstadt 4 Measurement procedure 4.1 Instrument The properties of the submitted newsprint samples were measured at IfP, Darmstadt under the guidance of Dr. H. Praast, who was responsible for the calibration of the instrument. Before the measurements were made, Dr. J. A. Bristow, convenor of the ISO/TC6/WG3 working group, undertook to check the instrument and the operating conditions. It had been emphasised at the first meeting of the working group that care was to be taken to ensure that measurements were made on a properly calibrated instrument in good working order. There were some doubts about the first instrument studied, particularly with regard to a directionality in the measurement of newsprint, and consequently an L&W Elrepho instrument was obtained on loan from AB Lorentzen & Wettre. In order to check the state of the instrument, Dr. J. A. Bristow had several series of samples which he had previously measured in a variety of different instruments, in particular the reference L&W Elrepho instrument placed at STFI. These samples were then measured in the L&W Elrepho on loan to IfP, in co-operation with Dr. H. Praast. The results showed a very good agreement between the two instruments and confirmed that the calibration and UV-adjustment had been satisfactorily carried out, and it was therefore judged that the instrument available at IfP was acceptable for the measurement of the samples collected by Ifra. For the purposes of this study, a non-fluorescent reference standard (IR3 type D) and a fluorescent standard for both D65- and C-illuminants (IR3 type FB) from STFI were used to calibrate and adjust the instrument for measurements under both UV(C) and UV(D65) conditions. 4.2 Measurements For each sample, a pad of 20 sheets was prepared. The first sheet on each side was for protection only. Measurements were made on each side of each of the next five sheets, by placing the pad of sheets in the instrument so that the machine direction (MD) of the paper was directed towards the rear of the L&W Elrepho instrument. The data for the two sides were averaged in the report prepared by IfP. Measurements were made under several sets of conditions: > all the samples were measured under UV(C) conditions > all the samples were measured under UV(D65) conditions > all the samples were measured under UVex(420) conditions. 4.3 Calculated parameters The following optical properties were calculated in accordance with the ISO standards, ISO 5631 (colour) and ISO 2470 (ISO brightness) for C/2 conditions: > Tristimulus values X, Y, Z > Chromaticity coordinates x, y > CIELAB colour coordinates L*, a*, b* > ISO brightness R457.
2004 Ifra, Darmstadt 5 Results 07 5 Results The primary results are summarised in two diagrams in which the C/2 data for all the samples are plotted in the CIELAB colour space. Figure 1a shows the chromatic variations in an a*b* diagram, and Figure 1b shows the lightness variations in an L*b* diagram. It is evident in Figure 1a that the standard newsprint grades form a cluster within a fairly small region, whereas the improved newsprint grades are much more widely scattered. In addition, it is evident in Figure 1b that the main difference between the standard and improved grades is that the latter have significantly higher L*-values. In view of this scatter, the Ifra Material Management Committee finds it impossible, in the present situation, to define any standard improved newsprint grade. The data show that the two VAC papers are significantly bluer and redder than the other papers, but that they have L*-values comparable with the improved newsprint grades. 5.1 Standard newsprint 5.1.1 Newsshade 2003 Table 2 shows the mean, maximum and minimum values and the standard deviations of the various parameters for the standard newsprint grades, together with the corresponding data obtained in the 1994 study. This table clearly shows that the various coordinates have changed very little since the previous investigation, but that both the L* and b* values are slightly lower. There may be several reasons for this change, but here it is not really possible to say to what extent this represents a real visual difference in shade and to what extent it is a difference created by different measurement conditions. Although the measurements were in both cases made in accordance with ISO 2469, several changes have nevertheless taken place during these seven years, and there are quite a lot of measurement differences between these two occasions. Firstly, there is a slight but significant difference between the 2001 and the 1994 calibration levels. In 1997, STFI announced a slight change in the absolute level of the photometric calibration values assigned to their non-fluorescent reference standards, which meant that the photometric scale was raised by about 0.2 reflectance factor units. 8 6 4 2 0 CIELAB b* standard news improved news VAC 2 illuminant/observer C/2 o 4 8 6 4 2 0 2 4 CIELAB a* 92 90 88 86 84 CIELAB L* VAC improved news standard news 82 illuminant/observer C/2 o 80 4 2 0 2 4 6 8 CIELAB b* Figure 1: C/2 data for all the samples plotted in a*b* and L*b* diagrams Secondly, there is a difference in the manner in which the UV-level in the spectrophotometers was adjusted. In 2001, the UV-content in the illumination was adjusted to match the C-illuminant using a fluorescent reference standard supplied by STFI with an assigned ISO-brightness Parameter Mean Minimum Maximum Standard deviation 2001 1994 2001 1994 2001 1994 2001 1994 L* 83.4 84.0 80.7 82.8 84.7 86.1 0.7 0.7 a* -0.49-0.56-1.33-1.27 0.42 0.11 0.33 0.32 b* 4.97 5.98 2.51 4.37 6.39 7.52 0.71 0.83 x 0.3194 0.3213 0.3140 0.3182 0.3234 0.3248 0.0015 0.0018 y 0.3268 0.3289 0.3220 0.3253 0.3295 0.3317 0.0015 0.0017 Y 62.9 63.6 58.0 61.8 65.4 68.2 1.3 1.4 Table 2: Data for standard newsprint measured in 2001 and in 1994
08 5 Results 2004 Ifra, Darmstadt value. In 1994 it was not however possible to adjust the UV-content of the illumination to match the illuminant C, and the UV-content in that case was set to zero with the aid of the adjustable 395 nm filter in the instrument. This was at that time the closest match to UV(C) conditions available. Thirdly, and perhaps most important, there was a difference in the type of instrument used. In 1994, the measurements were made on an Elrepho 2000 instrument and in 2001, the more modern L&W Elrepho instrument was used. Here, it is important to note, for example, that in 2001, after fairly extensive tests, STFI published a memorandum stating that the modern instruments appear in general to give results for newsprint which are significantly lower than the value given by the Elrepho 2000 instrument. In fact, the data in Table 2 appear to reflect these various changes. It is reasonable to suppose that mills have maintained the characteristics of their products during the years, particularly with regard to the Y-value. If this is the case then the mean drop in Y-value from 63.6 to 62.9 is exactly the sort of change which the STFI report predicts when a change is made from the old Elrepho 2000 to one of the new instruments. Although at first glance there appears to be no great change in the data from 1994 to 2001, the difference between the 1994 and 2001 means is E=1.18, where the colour difference DE is calculated according to the expression: E = [ L* 2 + a* 2+ b* 2 ] [ 1 ] This means that the 2002 mean is in fact outside the 1994 specifications for which the tolerance was E<0.7. It is thus evident that there is ample justification for slightly modifying the specifications for a standard newsprint in order to accommodate the changes referred to here. The Ifra Material Management Committee, acting on the recommendation of the working group, has thus decided that the set of values given in Table 3 shall define the standard newsprint Newsshade 2003. It is emphasised that in order to enable printers to communicate with the papermakers regarding the grades of newsprint which they wish to order, these coordinates are given in terms of the requirements of ISO 2469 and related International Standards, i.e. measurement over an opaque pad of the paper in a d/0 instrument having a 150 mm sphere and a gloss trap, with calibration and UVadjustment to UV(C) conditions using reference standards provided by an ISO/TC6 authorised laboratory. In its work to develop ISO 2469 and related International Standards, ISO/TC6/WG3 has sought to reduce to a 3 Authorized laboratories 4 Internal report, ISO/TC6/WG3:N465 minimum the potential sources of variation between different laboratories by defining rigidly the geometrical characteristics of the instruments, the routines for calibration and the adjustment of the UV-content of the illumination. Nevertheless, some variation is unavoidable. There are now several different makes of instrument which conform to the requirements of ISO 2469 and ISO/TC6 has authorised five different laboratories as suppliers of IR3 calibration standards. The values given in Table 3 should not, however, be interpreted as applying to any particular instrument or to a calibration traceable to any particular laboratory, but rather as being the best available representation of an instrument-independent shade. In a recent study 4, Dr. J. A. Bristow measured four grades of newsprint in one of each of five different types of instrument from different manufacturers, all conforming to ISO 2469. He found a mean E=0.22 considering the deviation of each instrument from the grand mean. When the now obsolescent Elrepho 2000 was excluded from the analysis, the variation dropped to E=0.17. The different authorised laboratories conduct regular comparative tests to assess the variation between the IR3 calibration standards supplied by the different ALs. Their variations are within E=0.1. The modern instruments produced by any given manufacturer are in very good agreement. One study of several instruments of one particular manufacturer has indicated that their variation is also less than E=0.1. Together these sources of variation give an expected variation of E= (0.222 + 0.12 + 0.12) 0.3. This unavoidable variation due to the different instruments and to the different calibration procedures must be taken into account, together with the uncertainty due to the variation within the material being measured, when the defined tolerance of E=0.7 associated with Newsshade 2003 is being interpreted. 5.1.2 Marketshade 2001 The standard in Table 3 is a standard with a narrow tolerance describing a specific shade of newsprint. The data presented in Table 2 and in Figure 1 can, however, also be used to provide a statistical characterisation of Marketshade 2001. Table 4 presents such data. In addition to the data previously presented, this table shows not only the standard deviation of the L*, a* and b* -values, but also the standard deviation of the differences between the individual values and the mean L*, a*, b* values. It also shows the mean and maximum E-values, and the standard deviations of the E-values. To describe a Marketshade with limits wide enough to include all these samples thus requires a specification of E=3 about the mean. Figure 1b shows, however, that the test material includes several samples with extremely low L*-values (corresponding to Y<60). If these samples are considered to be non-typical and if they are excluded from the analysis, the changes in the mean coordinates are very small, but the value of Emax drops to Emax=1.6.
2004 Ifra, Darmstadt 5 Results 09 Measurement specifications according to ISO 2469 Geometry d/0, 150 mm sphere, gloss trap Calibration IR3 supplied by ISO/TC6 AL 3 UV-adjustment to match CIE illuminant C Backing opaque pad of newsprint Colorimetric specifications Illuminant/observer C/2 L* 83.4 * a* -0.3 b* 5 Tolerance E<0.7 * the corresponding Y-value is 62.9 Table 3: Newsshade 2003, standard newsprint Parameter All standard newsprints mean standard deviation X 61.5 1.3 Y 62.9 1.3 Z 68.1 1.5 R457 58.1 1.2 L* 83.4 0.7 a* -0.5 0.3 b* 5 0.7 L* 0 0.7 a* 0 0.3 b* 0 0.7 E 0.9 0.5 E max 2.8 Table 4: Characteristics of Marketshade 2001, (C/2 ) Total radiance factor 100 80 60 40 News 23c News 9b News 30b 20 400 500 600 700 Wavelength, nm Figure 2: Spectral curves for three newsprints UV(C) conditions In this context, it should however be borne in mind that the sample material embraces a wide range of different newsprint grades and that the spectral curves are widely different. Figure 2 shows spectral radiance factor 5 curves for two extreme grades and an intermediate grade. The upper curve is typical of a newsprint based on virgin pulp and the lower curve is typical of a newsprint based on recycled fibre. The remainder of the samples show curves ranging between the two extremes. A study of the data suggests that about 50% of the samples are based on 100% recycled fibre. It is important to emphasise that two papers having the same ISO brightness values can have totally different shades. For instance, a paper made of virgin fibres and a paper containing DIP 6 can have the same ISO brightness, but the second will be more bluish than the first. These three samples all have brightness values in the range of 57 58, but the colorimetric difference between the two extremes is DE=3.7. In addition, the differences between the curves indicate that there is an unavoidable metamerism problem. 5.2 Other measurement results 5.2.1 Effect of fluorescence An important feature of modern newsprint grades is that they possess some degree of fluorescence, due either to the incorporation of recycled fluorescent material or to the addition of so-called optical brighteners. Excitation of these brighteners by UV-radiation leads to a fluorescent emission in the blue region of the spectrum and this means that the paper appears whiter. In this study, an attempt has been made to assess the extent to which fluorescence contributes to the brightness and colour of Marketshade 2001. Such information can be derived from the data obtained under UV(C), UV(D65) and UVex(420) conditions. The data obtained under UV(C) and UV(D65) conditions give by definition the ISO brightness and D65 brightness values, and the UVex(420) conditions give the spectral data for the material when the fluorescence has been eliminated, which is interpreted as being the brightness which the substrate would have if it contained no optical brightener. There is, however, some difficulty in assessing the contribution of the fluorescence to the ISO brightness, since the difference between the values obtained under UV(C) conditions and under UVex(420) conditions for standard newsprints are often negative. Dr. J. A. Bristow, convenor of the ISO/TC6/WG3 working group says that this problem has recently been discussed within this working group, but that no recommendation has yet been made. He points out that there is diffi- 5 The term "radiance factor is used rather than reflectance factor since the samples may be slightly fluorescent (see section 5.2.1) 6 Deinked pulp
10 5 Results 2004 Ifra, Darmstadt culty in computing the effect of the optical brightness by simple subtraction since the two sets of data are not fully compatible. The problem is that, under UV(C) conditions, Total radiance factor 100 90 80 70 60 50 400 500 600 700 Wavelength, nm Figure 3: Spectral radiance curves for one newsprint and one VAC sample under UVex(420), UV(C) and UV(D65) conditionsa Total radiance factor 100 90 80 70 60 50 400 500 600 700 Wavelength, nm Figure 4: Spectral radiance curves for two improved newsprints under UVex(420) and UV(C) conditions, with a superimposed brightness filter function (arbitrary scale) the calculation is based on total radiance factor data from 400 to 700 nm whereas, under UVex(420) conditions, data is available only from 420 to 700 nm. With extremely fluorescent materials, this discrepancy can be disregarded, but with materials such as newsprint with a very low degree of fluorescence, it introduces a problem. In the present analysis of brightness differences, two approaches have therefore been taken. In the first method, Dr. Bristow has recommended that the different sets of data be made compatible by restricting the calculations to the 420-700 nm range even under UV(C) and UV(D65) conditions. In accordance with accepted practice, the weighting factors for 400, 410 and 420 nm have then been combined. There are however still some difficulties with negative values, which may be due to experimental error or perhaps to secondary photochromic effects. For this reason, a second approach proposed to the ISO/TC6/WG3 working group by Mr. Hans Ekblad (L&W) has been tested. This approach is to take the difference between the spectral radiance curves obtained under UV(D65) and UV(C) conditions as being a true indication of the real fluorescence effect. It is known that the fluorescence emission under UV(D65) conditions is roughly twice the emission under UV(C) conditions, so that the difference between the ISO brightness values under these two conditions is in fact a good measure of the fluorescence contribution to the ISO brightness. Such values have here been calculated. Table 5 presents a summary of the data. In the analysis, it has been judged that a sample does not contain a significant amount of FWA 7 if the brightness value under UV(D65) conditions is less than 0.1 unit higher than the value under UV(C) conditions. With this definition, 77% of the standard newsprints and 59% of the improved newsprints are judged to contain FWA. It is seen in the table that, although the mean brightness of the improved newsprint samples is about 10 units higher than the mean brightness of the standard newsprint samples, the contribution of the fluorescence effect to this increase is very standard improved VAC Number of samples 97 59 2 Mean ISO brightness 58.3 68.6 77.1 Mean fluorescence contribution 0.5 0.6 3.4 Percentage containing FWA 77 59 100 Mean fluorescence contribution in samples containing FWA 0.7 1.1 3.4 Max fluorescence contribution 1.6 * 3.7 3.6 * one non-typical sample showed a fluorescence contribution of 4.8 brightness units Table 5: ISO brightness of samples fluorescence contribution 7 Fluorescent whiting agent
2004 Ifra, Darmstadt 5 Results 11 small. Similarly, the higher brightness of the VAC samples is not due solely to the presence of FWA. In the CIELAB colour space, the main effect of FWA is to lower the b*-value. In an attempt to throw more light on the behaviour indicated in Table 5, however, the spectral data have been studied. It is difficult to select single samples which may be considered representative of these different groups of papers. The following figures show selected examples. Figure 3 shows examples of the fluorescence effect for one newsprint sample and one VAC sample. For each sample, the three curves show respectively, from the bottom, the total radiance factor for the UVex(420) mode, UV(C) conditions, and UV(D65) conditions. The bottom curve shows the reflectance factor in the absence of any fluorescence, while the differences between this curve and the other two curves show the increasing effect of fluorescence as the UV-content is increased. It is here apparent that the fluorescence effect under UV(D65) conditions is, as previously mentioned, roughly twice as large as that under UV(C) conditions. Figure 4 shows similar curves under UVex(420) and UV(C) conditions for two improved newsprint grades. In addition, the ISO brightness (R457) filter function is shown superimposed on these curves. There is obviously a wide variation between these two grades, but this figure also emphasises that the brightness value relates to only a narrow portion of the spectrum. It remains ignorant of changes in other regions of the spectrum. The curves in this and the previous figure show other significant features. The general level of the curve is indicative of the degree of bleaching and the difference between the curves in each pair show how much is achieved by the addition of FWA, but the dip in the upper curves of the two figures in the region of 550 600 nm is indicative of the addition of a blue dye. Such dyes have no effect on the brightness value, but they do increase the perception of whiteness. This suggests, therefore, that a proper assessment of what the papermaker is trying to achieve should include the determination of a whiteness value. CIE whiteness (C/2 ) 90 R A 2=0.90 80 70 60 50 whiteness limit 40 30 20 50 55 60 65 70 75 80 ISO brightness Figure 5: CIE whiteness (C/2 ) plotted against ISO brightness for all the samples Frequency, % 30 20 10 0 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 ISO brightness, % Figure 6: Frequency distribution of ISO brightness values 5.2.2 CIE whiteness (C/2 ) In this study, the data obtained under UV(C) conditions have been used to compute the CIE whiteness (C/2 ) value in accordance with the requirements of ISO 11476. This value is calculated using the equation W=Y+800(x n -x)+1700(y n -y) [ 2 ] 20 10 Frequency, % where x n and y n are the chromaticity coordinates of the C-illuminant. Figure 5 shows for all the samples, the CIE whiteness (C/2 ) values plotted against the ISO brightness values. There is a general correlation with quite a high coefficient of determination (R 2 = 0.9), but it is also clear that there are significant differences. For a given brightness value, the whiteness value may span over roughly 20 whiteness 0 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 CIE whiteness (C/2 ) Figure 7: Frequency distribution of CIE whiteness (C/2 ) values
12 5 Results / 6 Colour designations under D50 conditions 2004 Ifra, Darmstadt units. The CIE whiteness equation contains a corollary that the value must exceed 40 for the paper to be considered white. This limit is drawn in the figure. In general, it can be said that the standard newsprint grades are below this limit and cannot thus be considered white, whereas the improved grades are placed slightly above this limit. Whiteness values of about 70 are not however usually considered to indicate a white paper. Figure 6 shows the frequency distribution of the ISO brightness values. Here it is evident that the standard newsprint grades have a brightness value in the region of 58, but that it is not possible to identify any value typical of the improved newsprint grades. As a complement to the discussion about CIE whiteness, Figure 7 shows for comparison the frequency distribution of the CIE whiteness values. Here, it can be seen that the standard newsprint grades have a whiteness value in the region of 35 40, which is below the whiteness limit, and that it is not possible to identify any region that can be considered to be characteristic of the improved newsprint grades. 6 Colour designations under D50 conditions 6.1 Colour designations under CIE illuminant D50 conditions In order to meet the requirements of the graphic arts industry, an attempt has here been made to transform these data to approximate D50/2 conditions, since the D50/2 illuminant/observer combination is recommended in the process standards developed by ISO/TC130, which is the Technical Committee responsible for the graphic arts sector (cf. ISO 13655). It is not yet possible to adjust the illumination in the spectrophotometers used in the paper mills to match this illuminant with regard to its UV-content but, in order to obtain reasonable values, it was here assumed that the fluorescence excitation under D50 conditions is approximately half way between the fluorescence excitations corresponding to the C and D65 illuminant conditions. The mean spectral radiance factors R( i ) D50 of the samples under D50 conditions were therefore estimated according to the formula: R( i ) D50 = [R( i ) C + R( i ) D65 ]/2 [ 3 ] where R( i ) C is the radiance factor for i under UV(C) conditions and R( i ) D65 is the corresponding radiance factor under UV(D65) conditions. The tristimulus values and CIELAB coordinates for D50/2 conditions have then been computed from the spectral data using the appropriate weighting functions. The values obtained are reasonably reliable, but it should be remembered that they still apply to opaque pads of paper and d/0 measurement conditions. The calculated mean values for standard newsprint corresponding to the data given in Table 4 for Marketshade 2001 standard newsprint are given in Table 9. The main difference between the values in the two tables lies in the a*-value. The redder characteristics of the D50 illuminant mean that the paper has a significantly redder a* coordinate under this illuminant condition (0.8, in contrast to -0.5 under C-illuminant conditions), and the paper is thus expected to appear redder under D50 than under C illumination. 6.2 Colour designations under D50 conditions over a black backing ISO 13655 which is concerned with conditions in a printing house, recommends that optical measurements be made on a single sheet over a black backing. In the present study, measurements have however only been made over an opaque pad of paper. The probable values over a black backing have, however, here been assessed by assuming that the opacity of the papers is 93% at all wavelengths. The spectral values have therefore been multiplied by the factor 0.93 and the tristimulus and CIELAB values calculated as before. The results are presented in Table 10. In general, this is a linear transformation which means that with respect to the values obtained over an opaque pad, the tristimulus values are simply multiplied by 0.93, and the chromaticity coordinates remain unchanged. The result of the non-linear transformation to CIELAB coordinates is that the a* and b* values remain unchanged and that the L* value is reduced by a factor of about 0.97.
2004 Ifra, Darmstadt 6 Colour designations under D50 conditions 13 It should be noted that in this calculation it has been assumed that the mean opacity of all the samples is 93%. Although there is a natural tendency for papers of lower grammage to have a lower opacity, this is not necessarily the case, since papermakers strive to maintain opacity even when there is a demand for a lower grammage. It has also been assumed that the opacity is the same at all wavelengths. In fact, the opacity may be slightly higher at shorter wavelengths, and this means that a paper seen over a black background may appear slightly less reddish. 6.3 The requirements of ISO 12647-3 The ISO/TC6 standards ISO 2469 etc and the ISO/TC130 standard 13655 are measurement standards specifying the conditions under which measurements are to be made and how certain parameters are to be calculated. ISO/TC130 is, however, also engaged in the development of process standards, and one such standard relates specifically to cold-set printing on newsprint is ISO 12647-3. In order to define the scope and region of applicability of the first version of this standard published in 1998, reference was made to the data relating to Newsshade 1995. In a future revision of this standard, reference should be made to this Ifra report relating to Marketshade 2001. The ISO/TC130 working group faces an unfortunate dilemma, since it needs to define the applicability of the standard both in terms of the measurement parameters familiar to and available to the printer in the press room and also in terms of the measurement parameters which are used by the papermaker to define the paper when it is bought and sold. The present report can only provide reliable data relating to the latter situation, although attempts have here been made to assess the significance of a change to controlled UV(D50) conditions in a d/0 geometry for standard newsprint grades over either an opaque pad or a black backing. As a complement to the calculations, a number of 45 g/m 2 samples that were closest to the target values were selected and measured under the printer s conditions in accordance with ISO 13655 and ISO 12647-1 (45 /0 or 0 /45 geometry, black backing D50-illuminant, 2 observer), using a 45 /0 GretagMacbeth SpectroEye instrument. The average values obtained are given in Table 11, with the corresponding values reproduced from Table 10 for comparison. The values are fairly close, although there are significant differences. In view of these discrepancies, it is recommended that the next revision of the standard (pre-release ISO/DIS 12647-3.2, expected at January 2004 and ISO 12647-3.2, expected at the end of 2004) should include the values for the standard Marketshade 2001 newsprint listed in Table 12. The value for standard Marketshade 2001 newsprint should be given a tolerance. For the maximum applicability of ISO 12647-3, a value of DE=3 seems desirable. Parameter All standard newsprints mean standard deviation X 61.0 1.4 Y 62.9 1.4 Z 49.2 1.1 R457 58.2 1.4 L* 83.4 0.7 a* 0.8 0.4 b* 3.0 0.7 L* 0 0.7 a* 0 0.4 b* 0 0.7 E 0.9 0.6 E max 3.0 Parameter All standard newsprints mean standard deviation X 56.8 1.3 Y 58.6 1.3 Z 45.8 1.0 R457 54.2 1.3 L* 81.1 0.7 a* 0.7 0.4 b* 3.0 0.7 L* 0 0.7 a* 0 0.4 b* 0 0.6 E 0.9 0.6 E max 3.0 Table 9: Characteristics of Marketshade 2001 D50/2 (measurements on opaque pads, d/0 geometry) Table 10: Characteristics of Marketshade 2001 D50/2 (calculated values over a theoretical black backing, d/0 geometry) L* a* b* 82.9-0.3 2.7 81.1 0.7 3.0 Table 11: D50/Black backing measurement values for standard newsprint L* a* b* 82 0 3 Table 12: D50/Black backing values for standard Marketshade 2001 newsprint
14 7 Conclusion 2004 Ifra, Darmstadt 7 Conclusion This study has been based on tests on 158 different samples of standard and improved newsprints, where each test has involved three sets of ten independent measurements. In order to provide a firm base for communication between the printer and the papermaker, the tests have been carried out in accordance with the International Standards developed by ISO/TC6 for the measurement of paper. The measurements were made in an L&W Elrepho instrument with a d/0 geometry calibrated in accordance with ISO 2469 and with UV-adjustment according to ISO 2470 and ISO 11475, and the results were expressed in terms of the C/2 illuminant/observer combination. In order to provide data of interest to the printer in the press room, attempts have been made to indicate the equivalent values over a black background under D50/2 conditions. The analysis of the data has provided statistical information describing Marketshade 2001, i.e. the range of different standard newsprint grades on the market in 2001 when the survey was initiated and the samples collected. The data have also provided a background against which it has been possible to define with narrow tolerances standard Newsshade 2003 of standard newsprint. The defined values are considered to be instrument-independent. The significance of these tolerances in relation to unavoidable sources of uncertainty among different instruments and different calibrations is discussed. Information concerning the range of improved grades available on the market is included in the statistical survey, but it has not been possible to identify and standardise any particular improved newsprint grade. It is emphasised that newsprint from different suppliers manufactured to this standard will have the same appearance only if the requirements of instrument geometry, traceable calibration and UV-adjustment are strictly adhered to. Even when this is done, problems of metamerism may still arise if the proportion of recycled fibre and the amount of FWA differ among the different grades. Measurement specifications according to ISO 2469 Geometry d/0, 150 mm sphere, gloss trap Calibration IR3 supplied by ISO/TC6 AL 3 UV-adjustment to match CIE illuminant C Backing opaque pad of newsprint Colorimetric specifications Illuminant/observer C/2 L* 83.4 * a* -0.3 b* 5 Tolerance E<0.7 * the corresponding Y-value is 62.9 ** Unless otherwise agreed between the parties, the tolerance is defined in relation to the mean value of material which has been sampled in accordance with the requirements of ISO 186, where measurements are made in accordance with the requirements of ISO 2469 and associated International Standards as indicated in this report, and due consideration is given to the uncertainty associated with the result of the measurements.
2004 Ifra, Darmstadt 8 References 15 8 References Ifra Special Report 1.11 Newsshade 95: a new proposed Ifra standard for newsprint shade and its measurement, 1996 Ifra Special Report 1.20 Value Added Coldset, 2001 ISO 186 Paper and board Sampling to determine average quality, 2002 ISO 2469 Paper, boards and pulps Measurement of diffuse reflectance factor, 1994 ISO 2470 Paper, boards and pulps Measurement of diffuse blue reflectance factor (ISO brightness), 1999 ISO 5631 Paper and board Determination of colour (C/2 degrees), Diffuse reflectance method, 2000 ISO 11475 Paper and board Determination of CIE-whiteness, D65/10 degrees (outdoor daylight), 1999 ISO 11476 Paper and board Determination of CIE-whiteness,, C/2 degrees (indoor illumination conditions), 2000 ISO 12647-1 Graphic technology Process control for the manufacture of half-tone colour separations, proofs and production prints, Part 1: Parameters and measurement methods, 1996 ISO 12647-3 Graphic technology Process control for the manufacture of half-tone colour separations, proofs and production prints, Part 3: Coldset offset lithography and letterpress on newsprint, 1998 ISO 13655 Graphic technology Spectral measurement and colorimetric computation for graphic arts images, 1996
Other Ifra Special Reports dealing with: 1 Materials (as of April 2004) 1.1 The Fracture Resistance of Newsprint A Joint Research Project of IFRA, SPPP and CTP 1.2 The bar code on the reel wrapping What possibilities does it offer to newspapers? 1.3 Deinkability of flexo printed newspapers 1.4 New Developments in Newsprint Warehouses 1.5 Vegetable oil based Newsinks and their Printability Properties 1.6 Paper recycling and deinking What is their importance for the newspaper industry? 1.7 Correlations between Ink Measuring Methods and Printing Results 1.8 Newspaper reel core and chuck specifications 1.9 Improvement of the Readability of Bar Codes on the Labels of the Newsprint Reels 1.10 The runnability and the printability of recycled newsprint 1.11 Newsshade 95: a new proposed Ifra standard for newsprint shade and its measurement 1.12 Why are soybean-based newsinks so successful in the U.S.A. 1.13 Introduction to the Problem of Copy Curl in Newspapers 1.14 Newsprint Waste Control: A Study of Leading U.S. and European Newspapers 1.15 The influence of the quality of raw water in damping solutions on newspaper printing 1.16 Runnability and Printability of Newsprint 1.17 EDI in newsprint acquisition 1.18 The performance of newsprint in newspaper production 1.19 The phenomenon of linting in newsprint printing 1.20 Value added coldset All topics of Ifra Special Reports 1 Materials 2 Pre-Press 3 Press 4 Mailroom and Distribution 5 Communication 6 General Should you wish to receive one or several copies of these Ifra Special Reports, contact: Ifra Washingtonplatz 64287 Darmstadt Germany Phone +49.6151.733-762 Fax +49.6151.733-800 www.ifra.com