ISO 12640-4:2011

INTERNATIONAL ISO
STANDARD 12640-4
First edition
2011-05-15

Graphic technology — Prepress digital
data exchange —
Part 4:
Wide gamut display-referred standard
colour image data
[Adobe RGB (1998)/SCID]
Technologie graphique — Échange de données numériques de
préimpression —
Partie 4: Données d'image standard montrées en référence par gamme
large de couleur [Adobe RGB (1998)/SCID]




Reference number
ISO 12640-4:2011(E)
©
ISO 2011

---------------------- Page: 1 ----------------------
ISO 12640-4:2011(E)

COPYRIGHT PROTECTED DOCUMENT


©  ISO 2011
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Published in Switzerland

ii © ISO 2011 – All rights reserved

---------------------- Page: 2 ----------------------
ISO 12640-4:2011(E)
Contents Page
Foreword .iv
Introduction.v
1 Scope.1
2 Normative references.1
3 Terms and definitions .1
4 Requirements.2
5 Data set characteristics .2
5.1 General .2
5.2 Data set definition .2
5.3 Natural images.3
5.4 Synthetic images .5
6 Electronic data.9
Annex A (normative) Guidance for use of digital data.10
Annex B (normative) Check-sum data.12
Annex C (informative) Typical TIFF/IT file header used for image files .13
Annex D (informative) Label text insertion.15
Annex E (informative) Histogram and gamut plots .17
Bibliography.25

© ISO 2011 – All rights reserved iii

---------------------- Page: 3 ----------------------
ISO 12640-4:2011(E)
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 12640-4 was prepared by Technical Committee ISO/TC 130, Graphic technology.
ISO 12640 consists of the following parts, under the general title Graphic technology — Prepress digital data
exchange:
⎯ Part 1: CMYK standard colour image data (CMYK/SCID)
⎯ Part 2: XYZ/sRGB encoded standard colour image data (XYZ/SCID)
⎯ Part 3: CIELAB standard colour image data (CIELAB/SCID)
⎯ Part 4: Wide gamut display-referred standard colour image data [Adobe RGB (1998)/SCID]
⎯ Part 5: Scene-referred standard colour image data (RIMM/SCID)

iv © ISO 2011 – All rights reserved

---------------------- Page: 4 ----------------------
ISO 12640-4:2011(E)
Introduction
0.1 Need for standard digital test images
Standard test images provide a set of data that can be used for any of the following tasks:
⎯ evaluating the colour reproduction of imaging systems;
⎯ evaluating colour image output devices;
⎯ evaluating the effect of image processing algorithms applied to the images;
⎯ evaluating the coding technologies necessary for the storage and transmission of high-definition image
data, etc.
Because they exist as standard, well-defined image data sets, typical of the high quality image content
commonly encountered, standard test images enable users to be confident that the images will produce good
quality reproductions, if properly rendered, and that they provide a reasonable test of the evaluation task being
undertaken. No limited set of images can fully test any system but the sets provided give as reasonable a test
as can be expected from a limited image set. Furthermore, the existence of a standard set enables users in
different locations to produce comparisons without the need to exchange images prior to reproduction.
However, different applications require that the standard image data be provided in different image states
using different image encodings (see ISO 22028-1). The user needs to select those appropriate to the
evaluation task being undertaken. Whilst transformation of the image data to another image state is always
possible, there is, in general, no agreement amongst experts as to how this is best done. Thus, it has been
considered preferable to provide data in different image states in the various parts of ISO 12640.

ISO 12640-1 provides a set of 8-bits-per-channel data that is defined in terms of CMYK dot percentages. The
colours resulting from reproduction of CMYK data are strictly defined only at the time of printing, and as such
the data are only applicable to evaluation of CMYK printing applications. Transformations to other image
states and colour encodings might not be well defined. In fact, the data might not even be useful for CMYK
printing processes different from those typically found in traditional graphic arts applications, as the image
data are defined to produce “pleasing” images when reproduced on systems using “typical” inks and
producing “typical” tone value rendering. Printing systems that use inks of a distinctly different colour, or
produce a very different tone value rendering, will not reproduce them as pleasing images without a well-
defined colour transformation. Moreover, with a bit depth of only 8 bits per channel any colour transformation
employed might well introduce artefacts.
ISO 12640-2 provides a set of test image data encoded both as XYZ values with each channel scaled to the
range 0 to 65 535, and as sRGB (defined in IEC 61966-2-1), with a bit depth of 8 bits per channel. (The higher
bit depth for the XYZ encoding is necessary because of the perceptual non-uniformity of the linear colour
space.) Both sets of data are optimized for viewing on a reference sRGB CRT display in the reference sRGB
viewing environment, and relative to CIE standard illuminant D65 for which the XYZ tristimulus values were
computed prior to scaling. The images are mainly designed to be used on systems utilizing sRGB as the
reference encoding, and as such are mainly applicable to the consumer market and those systems for which
the colour monitor is the “hub” device. Although such systems are used for some applications in the graphic
arts industry, sRGB is by no means the most common image encoding. Furthermore, a particular drawback is
the fact that the sRGB colour gamut is quite different in shape to the colour gamut of typical offset printing.
This difference can necessitate fairly aggressive colour re-rendering to produce optimal prints from sRGB
image data.

ISO 12640-3 provides a set of test image data with a large reflection medium colour gamut, illuminated using
illuminant D50. The bit depth of the natural images is 16 bits per channel, while the colour charts and
vignettes are 8 bits per channel. In order to be useful for applications where large, print-referred output
gamuts are encountered, common in graphic technology and photography, it was felt that it would be
desirable to produce an image set in which some colours are permitted to be encoded close to the boundary
© ISO 2011 – All rights reserved v

---------------------- Page: 5 ----------------------
ISO 12640-4:2011(E)
of the full colour gamut attained with surface colours. Furthermore, from the perspective of colour
management, it is advantageous if the images are referenced to illuminant D50, which is the predominant
reference illuminant used in graphic arts and photography, for both viewing and measurement. For this reason,
it has also become the predominant reference illuminant for most colour management applications.
This part of ISO 12640 provides a set of wide gamut test image data encoded as Adobe RGB with a bit depth
of 16 bits per channel. These data are optimized for viewing on a reference Adobe RGB display in the
reference Adobe RGB viewing environment [defined in the Adobe RGB (1998) Color Image Encoding
specification]. The preferred rendering of these images to any media, other than the reference Adobe RGB
display in the reference Adobe RGB viewing environment, is dependent on the media and viewing
environment used. Therefore, no colorimetry associated with reproduction on any other media is provided.
The images are mainly designed to be used on systems utilizing Adobe RGB as the reference encoding, and
as such are mainly applicable to the professional market and those systems for which the wide gamut colour
monitor is the “hub” device. Such workflows are popular among professional photographers, and are
increasingly used in the graphic arts. The Adobe RGB reference display colour gamut is closer to typical offset
printing gamuts than the sRGB reference display colour gamut. Adobe RGB encoded images generally
require much less aggressive colour re-rendering going to print than sRGB encoded images, although this
difference can necessitate colour re-rendering between Adobe RGB images and sRGB images. The purpose
of this part of ISO 12640 is therefore to provide a test image data set with a larger colour gamut than sRGB,
related to the Adobe RGB wide gamut display-referred colour space. The bit depth of the natural images and
synthetic images is 16 bits per channel.
The possible wide gamut colour encoding choices considered were Adobe RGB, opRGB (IEC 61966-2-5) and
ROMM RGB (ISO 22028-2). For this part of ISO 12640 we want the images well-colour-rendered to a well-
defined large gamut reference display. For this reason, Adobe RGB was preferred over the other two choices.
With opRGB the completeness of the colour rendering is left more ambiguous, i.e. it is not as clearly output-
referred, and the reference medium and viewing conditions are also slightly different. ROMM RGB
(ISO 22028-2) is clearly output-referred, but the reference medium is a virtual reflection print (the ICC
perceptual reference medium).
0.2 Characteristics of the test images
The performance of any colour reproduction system will normally be evaluated both subjectively (by viewing
the final output image) and objectively (by measurement of control elements). This requirement dictates that
the test images include both natural scenes (pictures) and synthetic images (colour charts and colour
vignettes). Because the results of subjective image evaluation are strongly affected by the image content, it
was important to ensure that the natural images were of high quality and contained diverse subject matter.
However, by requiring the images to look natural, it is difficult within a single, relatively small, sample set to
produce elements in the scene that contain the subtle colour differences required in such test images, that
cover the full reference colour gamut defined. Thus, while most images contain colours that extend to the
gamut boundary this is often only for a limited range of hues in each image. The full reference colour gamut
can only be explored by utilizing the synthetic colour chart.
A survey was conducted of all TC 130 member countries to identify desirable image content and to solicit
submission of suitable images for consideration. The image set that resulted consists of 14 natural images, a
colour chart and a series of colour vignettes. The natural images include flesh tones, images with detail in the
extreme highlights or shadows, neutral colours, brown and wood tone colours which are often difficult to
reproduce, memory colours, complicated geometric shapes, fine detail, and highlight and shadow vignettes.
The colour chart and colour vignette show the colour gamut of this wide gamut display-referred colour space.
0.3 File format of the digital test images
All of the images consist of pixel interleaved data (R then G then B) with the data origin at the upper left of the
image, as viewed naturally, and organized by rows. These data are included as individual files within this part
of ISO 12640. The image file format is as specified in ISO 12639:2004 (TIFF/IT). The images can be imported
and manipulated as necessary by a wide variety of imaging software tools and platforms in general use in the
industry. See Annex C for details of the TIFF header.

vi © ISO 2011 – All rights reserved

---------------------- Page: 6 ----------------------
INTERNATIONAL STANDARD ISO 12640-4:2011(E)

Graphic technology — Prepress digital data exchange —
Part 4:
Wide gamut display-referred standard colour image data
[Adobe RGB (1998)/SCID]
IMPORTANT — The electronic file of this document contains colours which are considered to be
useful for the correct understanding of the document. Users should therefore consider printing this
document using a colour printer.
1 Scope
This part of ISO 12640 specifies a set of standard wide gamut display-referred colour images [encoded as
16-bit Adobe RGB (1998) digital data] that can be used for the evaluation of changes in image quality during
coding, image processing (including colour re-rendering and colour space transformations, compression and
decompression), displaying on a colour monitor and printing. These images can be used for research, testing
and assessing of output systems such as printers, colour management systems and colour profiles.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
ISO 12639:2004, Graphic technology — Prepress digital data exchange — Tag image file format for image
technology (TIFF/IT)
Adobe RGB (1998) Color Image Encoding, Version 2005-05, May 2005. Available at:

3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
3.1
Adobe RGB
three-component colour image encoding defined in Adobe RGB (1998) Color Image Encoding
3.2
colour gamut
solid in a colour space, consisting of all those colours that are present in a specific scene, artwork, photograph,
photomechanical or other reproduction; or are capable of being created using a particular output device or
medium
[ISO 12231]
© ISO 2011 – All rights reserved 1

---------------------- Page: 7 ----------------------
ISO 12640-4:2011(E)
3.3
colour sequence
order in which the colours are stored in an image data file
3.4
orientation
origin and direction of the first line of data, with respect to the image content as viewed by the end user
NOTE The codes used to specify orientation are listed in ISO 12639:2004.
3.5
pixel
smallest discrete picture element in a digital image file
3.6
pixel interleaved
colour data organized such that the RGB colour space values for one pixel are followed by the same
sequence of colour values for the next pixel
NOTE The specific order of colour components is determined by the ColourSequence tag as defined in
ISO 12639:2004. Other forms of colour data interleaving are line and plane.
4 Requirements
This part of ISO 12640 provides a set of images contained in 16 image data files which form an integral part of
this part of ISO 12640. Their file names are listed in Tables 1 and 2. The colour image data shall be encoded
as defined in Adobe RGB (1998) Color Image Encoding (hereafter referred to as Adobe RGB) using 16 bits
per channel and 48 bits per colour. The image characteristics of these data are described in Clause 5, and the
electronic data structure is described in Clause 6.
The procedures and guidelines for use of the image data files are given in Annex A. The image data integrity,
excluding any headers, shall be checked using the check-sum procedure outlined in Annex B. Typical TIFF/IT
file headers used for image files are described in Annex C. Label text insertion is described in Annex D. The
histogram and gamut plots for the image data files are shown in Annex E.
5 Data set characteristics
5.1 General
The orientation of the image data is defined in accordance with ISO 12639, where a value of “1” in TAG 274
indicates that the data are to be loaded from top left, horizontally; the 0th row represents the visual top of the
image and 0th column represents the visual left-hand side. The image data are pixel interleaved in the colour
sequence of R then G then B (16 bits per channel) for the natural images and synthetic images.
5.2 Data set definition
The set of standard colour image data consists of 14 natural (photographed) images and two synthetic images
created digitally on a computer. The synthetic images consist of one colour chart with various patches, and
one colour vignette. The natural images are identified as N1 to N14, respectively, and each of them also has a
descriptive name derived from the picture content (e.g. Crayons). The synthetic images are identified as S1
and S2.
The label “ISO 12640-4 RGB” is inserted in each image. The co-ordinates of the text insertion are provided in
Annex D.
2 © ISO 2011 – All rights reserved

---------------------- Page: 8 ----------------------
ISO 12640-4:2011(E)
NOTE The image set defined in this part of ISO 12640 is based on the Adobe RGB reference display gamut. Image
sets contained in other parts of ISO 12640 are based on different reference media and can be more suitable for use in
evaluations where the other reference media are more relevant.
5.3 Natural images
The characteristics and typical usage of the natural images are shown in Table 1. The descriptive names of
these images are given following the identification code. Figure 1 shows reduced size reproductions of the
natural images.
The 14 natural images have the following characteristics:
Resolution: 16 pixels/mm
Colour values: Adobe RGB data consisting of three 16-bit values
File format: ISO 12639:2004 (TIFF/IT)
Label on image: ISO 12640-4 RGB
Image data orientation: Load from top left, horizontally
Table 1 — Natural images
Name Aspect, image size Characteristics and typical usage
N1 Crayons Horizontal, Picture of crayons with high saturation colours; useful for checking edge-
4 096 × 3 072 pixels of-gamut reproduction
N2 Flowers Vertical, Useful for assessing tonal reproduction of highlight tones and saturated
3 072 × 4 096 pixels reds
N3 Yarn Horizontal, Image of yarn, wool and thread suitable for evaluating the colour gamut of
4 096 × 3 072 pixels devices, texture and fine detail reproduction
N4 Fishing Vertical, Fishing goods with fine detail, suitable for evaluating image sharpness and
reproduction of detail
3 072 × 4 096 pixels
N5 Vases Horizontal, Picture of transparent and semi-transparent vases, suitable for evaluating
4 080 × 3 072 pixels the reproduction of smooth highlight tones
N6 Leaves Horizontal, Useful in evaluating the reproduction of subtle tonal variation in the leaves
4 096 × 3 072 pixels and of shadow detail in the dark brown of the trunks of the trees
N7 Borabora Horizontal, Landscape image; suitable for the evaluation of the reproduction of deep
blue and green colours with subtle tonal variation
4 124 × 3 024 pixels
N8 Sunflower Horizontal, Field of sunflowers with memory colours for sky, trees and grass; suitable
for evaluating the reproduction of natural scenes
3 040 × 2 014 pixels
N9 Bride Vertical, Close-up image to evaluate the reproduction of human skin tones
3 072 × 4 096 pixels
N10 Walkathon Vertical, Image of children in walking gear with bright balloons can be used to check
2 000 × 3 008 pixels the reproduction of images that include saturated colours and skin tones
N11 Spoon Horizontal, Image of silverware to evaluate the reproduction characteristics of highlight
tones and neutral colours
4 096 × 3 072 pixels
N12 Violin Vertical, Low-key image of a room scene containing miscellaneous objects to
evaluate dark colours, particularly browns
3 072 × 4 096 pixels
N13 Glass Horizontal, Image of glassware to evaluate the reproduction characteristics of highlight
4 096 × 3 072 pixels tones, shadow tones and neutral colours
N14 Beach Horizontal, Image of a sunny beach shot from shade of trees can be used to evaluate
3 040 × 2 014 pixels the reproduction of images having a high dynamic range

© ISO 2011 – All rights reserved 3

---------------------- Page: 9 ----------------------
ISO 12640-4:2011(E)

N1  Crayons N2  Flowers

N3  Yarns N4  Fishing

N5  Vases N6  Leaves

N7  Borabora N8  Sunflower
Figure 1 (continued)

4 © ISO 2011 – All rights reserved

---------------------- Page: 10 ----------------------
ISO 12640-4:2011(E)

N9  Bride N10  Walkathon

N11  Spoon N12  Violin

N13  Glass N14  Beach
Figure 1 — Reduced size reproductions of the natural images
5.4 Synthetic images
5.4.1 Synthetic image content
The synthetic images consist of a colour chart and a series of colour vignettes. Figure 2 shows reduced-size
reproductions of the synthetic images. The interleaving, colour sequence, colour values and orientation are
the same as for the natural images. The image sizes are shown in Table 2.
Table 2 — Synthetic images
Height Width
Name Aspect
pixels pixels
S1 Colour chart Landscape 1 332 2 736
S2 Colour vignettes Landscape 2 608 4 256
© ISO 2011 – All rights reserved 5

---------------------- Page: 11 ----------------------
ISO 12640-4:2011(E)

S1  Colour chart S2  Colour vignettes
Figure 2 — Reduced-size reproductions of the synthetic images
5.4.2 Colour chart
5.4.2.1 Design of colour chart
Image S1 (shown schematically in Figure 3) is a colour chart that consists of colour patches that are all
specified to be within the Adobe RGB reference display colour gamut. By using these patches, the fidelity of
colour reproduction of an image output device to the colorimetry of the original image file may be evaluated
objectively by measurement. Image S1 is encoded in 16-bit Adobe RGB. Each part of the chart has two
sections:
3
⎯ section containing 6 (i.e. 216) tertiary colour patches;
⎯ primary, secondary and tertiary grey colour section (77 patches in total).
The Adobe RGB image data encoded can be converted to viewer-observed image colorimetry using the
transforms specified in Adobe RGB. The complete transformation, through the normalized tristimulus values to
the viewer-observed tristimulus values, should be used.
NOTE When comparing the fidelity of a colour reproduction to that of an original, it is generally most appropriate to
compare viewer-observed colorimetric values. However, the fidelity of measured reproduction colorimetry to original image
colorimetry is not generally considered as indicative of the quality of the reproduction. To produce optimal quality, it is
frequently necessary to adjust the colorimetry of a reproduction to be different from that directly associated with the image
data in order to account for any differences between the Adobe RGB viewing conditions and the reproduction viewing
conditions, and because of differences between the Adobe RGB and reproduction medium colour gamuts.
6 © ISO 2011 – All rights reserved

---------------------- Page: 12 ----------------------
A B C D E F G H I J K L
ISO 12640-4:2011(E)
123456 789101112 131415161718 19202122232425
C1.1
C1.2 C1.3
C1.4 C1.5 C1.6
C2.1 C2.4 C2.7
C2.5
C2.2
C2.3 C2.6
3
6 section
Primary, secondary
and tertiary sections

Figure 3 — Colour chart (S1)
3
5.4.2.2 Generation of the content of the 6 colour section
3
The Adobe RGB image data for the 6 section of image S1 were obtained by the following procedure:
a) Determine step differences by uniformly dividing the black and white normalized (not viewer-observed)
range of L* (0 to 100) into six steps.
b) Convert each normalized L* value (assuming a* and b* values of 0) to 32-bit floating XYZ data normalized
to range from 0 to 1; that is the maximum XYZ values are 0,950 5, 1,000 0 and 1,089 1, respectively.
c) Convert the resulting six normalized XYZ values to linear RGB values using Equation (1).
⎡⎤R ⎡ 2,,04159−−0 56 501 0,344 73⎤⎡X⎤
⎢⎥ ⎢ ⎥⎢ ⎥
G,=−0 969 24 1,875 97 0,04156Y (1)
⎢⎥ ⎢ ⎥⎢ ⎥
B 0,,013 44 −0 118 36 1,015 17Z
⎣⎦ ⎣ ⎦⎣ ⎦

Obtain 216 RGB combinations by combining the 6 resultant linear RGB values in all combinations.
3
d) Obtain 16-bit Adobe RGB (1998) component values R', G' and B' corresponding to each of the 6 linear
RGB values using Equation (2).
1/2,199 218 75
⎫
′
RR=×Round (65 535 )
⎪
⎪
1/2,199 218 75
′
GG=×Round (65 535 ) (2)
⎬
⎪
1/2,199 218 75
′
BB=×Round (65 535 )
⎪
⎭
Between the 6 blocks (C.1.1 to C.1.6) the G value is altered as the block parameter. Within each block, the R
value is stepped along the horizontal direction, and the B value along the vertical direction.
NOTE This clause follows the procedure described in 4.3.4 of Adobe RGB (1998) Color Image Encoding.
© ISO 2011 – All rights reserved 7

---------------------- Page: 13 ----------------------
ISO 12640-4:2011(E)
5.4.2.3 Generation of the content of the primary, secondary and tertiary grey colour section
The Adobe RGB data for columns 19 to 25 were prepared using the following procedures:
a) Step a) of the procedure defined in 5.4.2.2 was repeated, except that the division of the range of L* (from
0 to 90) was changed from 6 to 10 steps, and one step of L*=5 was added to produce 11 steps altogether.
Then steps b) and c) defined in 5.4.2.2 were used to obtain the linear RGB values.
b) Suitable combinations of linear RGB values were defined to provide the primary (red, green, blue),
secondary (cyan, magenta, yellow) and tertiary (grey) colours.
c) By using the same procedures as those described in step d) of 5.4.2.2 the 16-bit Adobe RGB values were
computed for each linear RGB combination.
It should be noted that the reason for not including a 12th level in each of these scales is that the logical
choice would have been the white point (L*=100, a*=b*=0). But this step is common to all the colours and is
already included in the tertiary colour section.
These scales vary in the vertical direction in terms of RGB, and are arranged horizontally in the order R, G, B,
C, M, Y and grey.
5.4.3 Colour vignettes
5.4.3.1 Description of the colour vignettes
Image S2 is a set of colour vignettes in which the lightness continuously changes along the horizontal
direction. By using this pattern, it is possible to evaluate the tone reproduction characteristics, or the number
of reproducible tonal levels, which may be obtained with any output device. It is possible visually to judge the
effects of the important image processing tasks of tonal modification or data compression
...