ISO 17321-1:2012

INTERNATIONAL ISO
STANDARD 17321-1
Second edition
2012-11-01
Graphic technology and
photography — Colour
characterisation of digital still
cameras (DSCs) —
Part 1:
Stimuli, metrology and test procedures
Technologie graphique et photographie — Caractérisation de la
couleur des appareils photonumériques —
Partie 1: Stimuli, métrologie et modes opératoires d’essai
Reference number
©
ISO 2012
© ISO 2012
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 2012 – All rights reserved

Contents Page
Foreword .iv
Introduction .v
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 DSC colour characterization methods . 2
4.1 General . 2
4.2 Spectral sensitivity-based characterization — Method A . 3
4.3 Target-based characterization — Method B . 4
Annex A (informative) Recommended laboratory set-up for photographing a reflection colour
test target . 8
Annex B (informative) Digital still camera/sensitivity metamerism index (DSC/SMI).10
Annex C (informative) Characterization target considerations .16
Annex D (informative) Calculating natural scene element responses from spectral
characterization data . .23
Bibliography .26
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 17321-1 was prepared by Technical Committee ISO/TC 42, Photography, in collaboration with
ISO/TC 130, Graphic technology.
This second edition cancels and replaces the first edition (ISO 17321-1:2006), of which it constitutes a
minor revision with the following changes:
— in 4.3.3.4.2, the typographical error “senor image area” was corrected to “sensor image area”;
— in B.2.6, a broken link to tools for non-linear optimization has been updated.
ISO 17321 consists of the following parts, under the general title Graphic technology and photography —
Colour characterization of digital still cameras (DSCs):
— Part 1: Stimuli, metrology and test procedures
— Part 2: Considerations for determining scene analysis transforms [Technical Report]
iv © ISO 2012 – All rights reserved

Introduction
The spectral responses of the colour analysis channels of digital still cameras (DSCs) do not, in general,
match those of a typical human observer, such as defined by the CIE standard colorimetric observer. Nor
do the responses of different DSCs ordinarily match each other. In characterizing DSCs, it is therefore
necessary to take account of the DSC spectral sensitivities, illumination, and encoding colour space.
This part of ISO 17321 will begin to address these considerations. This part of ISO 17321 defines stimuli
(spectral illumination or a colour target), metrology and photographic test procedures for acquiring
DSC characterization data. It specifies test procedures for “scenes”, the most general picture taking
conditions where metameric colours and a range of illumination sources are encountered. It also
specifies test procedures for hardcopy “originals”, a more narrowly defined picture-taking condition in
which the illumination source and the colorants being imaged are pre-defined.
ISO 17321 will distinguish among several possible image representations in different colour encodings
as depicted in Figure 1 which shows the diagram of a generic image workflow for digital photography.
Figure 1 — Generic image workflow for digital photography
The DSC characterizations obtained using this part of ISO 17321 will be applicable to raw (sensor-
referred) DSC data. Two alternative methods are described for obtaining these characterization data.
Method A, the spectral method, uses spectral lights as stimuli for measuring the colour performance
of a DSC. Method B, the target method, involves the use of a physical colour test target under specific
lighting conditions to measure DSC colour performance. Annexes A to C recommend a laboratory set-up
for photographing reflection targets, provide target patch selection criteria, and provide a digital still
camera metamerism index.
Some operations (colour pixel reconstruction, flare removal, white balancing) can be performed without
disqualifying the DSC data as being raw. However, operations that render the image data so that they
become output-referred (ready to display or to print) generally do disqualify the data. With such cameras,
this standard can only be applied if the capability exists to extract or to regenerate raw data, e.g. by applying
the inverse of the rendering transform or by tapping the appropriate signals internal to the camera.
The technical experts who have developed this part of ISO 17321 recognize that a standard that could
be applied generally to any (not just raw) DSC output would be desirable. Such a standard is problematic
for DSCs that employ colour-rendering algorithms in order to produce output-referred image data.
For such DSCs, it would frequently be impossible to determine if colour analysis errors relative to the
scene or original captured were due to sensor image encoding errors or to proprietary colour rendering
algorithms. The only way to make this distinction is if the colour rendering used is well documented and
available, and the rendered data can be converted to un-rendered data by inverting the colour rendering.
This situation is unlikely to occur because one of the major differentiators in DSC performance is the
colour rendering. Sophisticated colour-rendering algorithms can be image dependent, and locally varying
within an image. This makes it extremely difficult to reliably determine the exact colour rendering used
by analysing captured test scenes.
The purpose of this part of ISO 17321 is both to assist in the characterization of DSCs for colour
management purposes and to assist camera manufacturers in the determination of the colour analysis
capabilities of DSCs that they are developing. This standard is applicable to any DSC intended for
photographic or graphic technology applications. However, for many users it is not practical to apply
this part of ISO 17321 to individual DSCs. Some of the measurements described in this part of ISO 17321
require complex, expensive measurement equipment. In the case of test targets that are commercially
produced, spectral as well as colorimetric measurement data would ideally accompany the target.
Those unfamiliar with this part of ISO 17321 are encouraged to read through the entire standard (in
particular the informative annexes) before proceeding with DSC characterization, in order to verify
appropriateness for their particular application. In some cases, the procedures described in the
[5]
multimedia standard, IEC 61966-9 might be more applicable.
It is proposed that other parts of ISO 17321 will be developed in the future to deal with other aspects of
the colour characterization of digital still cameras.
vi © ISO 2012 – All rights reserved

INTERNATIONAL STANDARD ISO 17321-1:2012(E)
Graphic technology and photography — Colour
characterisation of digital still cameras (DSCs) —
Part 1:
Stimuli, metrology and test procedures
1 Scope
This part of ISO 17321 specifies colour stimuli, metrology, and test procedures for the colour
characterization of a digital still camera (DSC) to be used for photography and graphic technology. Two
methods are provided, one using narrow spectral band illumination and the other using a spectrally
and colorimetrically calibrated target. Except for a specific set of permitted data operations, these DSC
data are raw.
This part of ISO 17321 does not specify the methods for deriving transformations from raw DSC data in
order to estimate scene colorimetry.
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 7589, Photography — Illuminants for sensitometry — Specifications for daylight, incandescent
tungsten and printer
ISO 13655, Graphic technology — Spectral measurement and colorimetric computation for graphic arts images
ISO 14524:2009, Photography — Electronic still-picture cameras — Methods for measuring opto-electronic
conversion functions (OECFs)
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
3.1
adopted white
spectral radiance distribution as seen by an image capture or measurement device and converted to
colour signals that are considered to be perfectly achromatic and to have an observer adaptive luminance
factor of unity; i.e. colour signals that are considered to correspond to a perfect white diffuser
[ISO 22028-1]
NOTE 1 The adopted white can vary within a scene.
NOTE 2 No assumptions can be made concerning the relation between the adapted or adopted white and
measurements of near perfectly reflecting diffusers in a scene, because measurements of such diffusers will
depend on the illumination and viewing geometry, and other elements in the scene that can affect perception.
3.2
digital still camera
DSC
device that incorporates an image sensor and that produces a digital signal representing a still picture
NOTE A digital still camera is typically a portable, hand-held device. The digital signal is usually recorded on
a removable memory, such as a solid-state memory card or magnetic disk.
3.3
opto-electronic conversion function
OECF
relationship between log of input levels and corresponding digital output levels for an opto-electronic
digital image capture system
NOTE If the input log exposure points are very finely spaced and the output noise is small compared to
the quantization interval, the OECF possibly has a step-like character. Such behaviour is an artefact of the
quantization process and needs to be removed by using an appropriate smoothing algorithm or by fitting a
smooth curve to the data.
3.4
raw DSC image data
image data produced by, or internal to, a DSC that has not been processed, except for A/D conversion and
the following optional steps:
— linearization,
— dark current/frame subtraction,
— shading and sensitivity (flat field) correction,
— flare removal,
— white balancing (e.g. so the adopted white produces equal RGB values or no chrominance),
— missing colour pixel reconstruction (without colour transformations)
3.5
spectrally non-selective
exhibiting reflective or transmissive characteristics that are constant over the wavelength range of interest
4 DSC colour characterization methods
4.1 General
Two methods are specified for obtaining raw DSC colour characterization data, a spectral method and
a target method. The method that is most applicable in any particular situation depends on a variety of
factors including, but not limited to, the following:
— the extent of one’s prior knowledge about the spectral content of the scenes or originals to be captured;
— the equipment available;
— the accuracy required.
The spectral method requires elaborate equipment in a laboratory environment, but can be used to
produce characterization data for samples with arbitrary spectral distributions. The target method is
suitable for studio and field use, but can only provide accurate characterization data to the extent that
the target spectral characteristics match those of the scene or original to be photographed.
2 © ISO 2012 – All rights reserved

4.2 Spectral sensitivity-based characterization — Method A
4.2.1 Equipment
4.2.1.1 General
Spectral sensitivity-based characterization measurements shall be obtained by using a light source and
monochromator to evenly illuminate a diffuse transmissive or reflective surface with electromagnetic
radiation (light) containing a limited range of wavelengths centred on selected wavelengths, as specified
in 4.2.3. Integrated relative radiance measurements of the illuminated surface shall be obtained for
each selected wavelength using a radiance or irradiance meter with a spectral sensitivity calibration
accurate to within 0,1 % and traceable to a national standards laboratory.
4.2.1.2 Light source
The light source shall output radiation where the power is a smooth function of the wavelength, such as
that obtained from a quartz-halogen source. Light sources that have strong emission lines shall not be used.
NOTE A fluorescent lamp is a typical light source with strong emission lines.
4.2.1.3 Monochromator spectral sampling and band pass
The bandpass of the illuminating instrument (monochrometer) shall be 5 nm or narrower. The sampling
interval shall not be greater than the bandpass. The monochromator should exhibit an approximately
triangular band pass, with the full width at half-maximum wavelength range approximately equal to the
sample spacing. The integrated radiance at all wavelengths more than 10 nm from the peak wavelength
on which the monochromator is set shall be less than 1/1 000, and should be less than 1/10 000, of the
integrated radiance within 10 nm of the peak radiance. Interference filters or a double monochromator
may be used to meet this requirement.
4.2.1.4 Illuminated surface
The illuminated surface should be the interior of an integrating sphere. It is recommended to obtain an
integrating sphere with three ports close together. A transmissive diffuser is placed over one port, and
illuminated by the monochromator. This produces an even illumination on the interior of the sphere.
The second port is for the DSC, and the third port is for the radiance or irradiance meter. Other evenly
illuminated surfaces may be used, but it is the responsibility of the user to ensure such surfaces do not
have characteristics that could influence the measurements. In all cases, stray light shall be prevented
from entering the integrating sphere or camera.
NOTE 1 This can be achieved by carefully enclosing the integrating sphere with the camera attached with an
opaque black fabric or plastic.
NOTE 2 The radiances produced for this measurement, and for the OECF measurement described in 4.2.3, need
to be comparable to those encountered in the normal operation of the digital camera.
4.2.2 Camera settings
Fixed exposure settings shall be selected to provide peak output levels between 50 % and 90 % of
saturation. Any automatic gain or adaptive tone reproduction (analog or digital) shall be disabled,
compression shall be minimized, and all user settings shall be recorded. White balancing (analog or
digital) shall be fixed, so that variations in white balance do not influence the measurements. Flash
should be disabled to reduce the possibility of stray light.
4.2.3 Capturing of raw images of the output of the monochromator
The procedure for capturing raw image data using Method A shall be as follows.
a) Use a monochromator to illuminate a diffuse transmitting or reflecting surface with light centred
on selected wavelengths so the illuminated area is large enough to fill the field of view of the DSC.
The radiance fall-off at the DSC focal plane should be even, constant as the monochromator peak
wavelength is changed, and circularly symmetric with the radiance at the edge no less than 70 % of
the radiance at the centre.
b) Use a radiance or irradiance meter to measure the relative radiance of the illuminated surface as a
function of wavelength.
c) Capture images of the illuminated surface at wavelengths ranging from 360 nm to at least 830 nm,
and preferably to 1 100 nm in 10 nm or smaller increments. The DSC shall be set up as described
in ISO 14524 for alternative focal plane OECF measurements. The images shall be captured with
the DSC lens and any filters used for general picture taking (such as an infrared blocking filter) in
place. The data output by each colour analysis channel of the DSC shall remain independent, i.e. not
be matrixed. The relative radiance of the surface shall also be recorded for each image. Where the
DSC under test can be shown to have essentially no sensitivity at wavelengths within the above
wavelength ranges, these ranges may be truncated appropriately.
d) Determine the alternative focal plane OECF of the DSC in accordance with ISO 14524, except that the
measurement may be performed at the peak sensitivity wavelength for each colour analysis channel.
4.2.4 Post-processing of the data
Use the inverse alternative focal plane OECF to linearize the raw DSC responses at each wavelength.
Average a 64 × 64 pixel block of values at the centre of each image to determine the linearized DSC
response at each wavelength.
4.2.5 Calculation of the relative spectral sensitivities of the DSC
Calculate the relative spectral sensitivities at each wavelength for each colour analysis channel by
dividing the linearized DSC response by the relative surface radiance. Optionally, DSC relative response
values for various scene spectral radiances may be calculated by taking the scalar product of the
spectral radiance and the spectral sensitivity vectors for each DSC colour channel (see Annex D for
more information).
Normalize the spectral sensitivities so the sum of the green channel sensitivities is unity. A different
channel may be normalized to unity if so reported.
NOTE If desired, the OECF of each channel, as measured in accordance with ISO 14524, can be used to
determine absolute spectral sensitivities from the relative spectral sensitivities.
4.2.6 Data reporting
The data shall be reported in tabular form, with the relative spectral sensitivity reported for each
channel at each selected wavelength.
4.3 Target-based characterization — Method B
4.3.1 General
The method to be used for the collection of target-based characterization data consists of imaging a
reflective or transmissive colour target of known spectral and colorimetric characteristics, under
4 © ISO 2012 – All rights reserved

specified illumination, recording the output of the DSC for each patch, and providing these data for
subsequent processing.
NOTE When target-based characterization is used, the resultant characterization data is only applicable for
similar geometric and spectral illumination characteristics.
4.3.2 Test target
The choice of test target to be used shall be a decision of the individual doing the characterization
and may be a commercially available target or a custom target developed for the purpose of digital
still camera characterization. Annex C provides a listing of some of the characteristics that should be
considered in developing an ideal target for DSC characterization.
Regardless of the target used it shall have available a tabulation of the spectral reflectance factor or
spectral transmittance factor for each patch. This data shall be from at least 380 nm to 730 nm at least
at every 10 nm and should be from 360 nm to 830 nm at least at every 10 nm. In addition, colorimetric
values for all of the colour patches should be included. Measurements and computation of colorimetric
parameters shall be in accordance with ISO 13655. The use of telephoto-spectrometer from the identical
position where the camera is set is preferred.
Where the DSC under test can be shown to have essentially no sensitivity at wavelengths within the
above wavelength ranges, the target measurement requirements may be truncated appropriately.
NOTE 1 Two commonly available commercial targets that many have used for this application are the traditional
24 patch ColorChecker and the 237 patch ColorChecker DC Digital Camera Color Reference Chart. (ColorChecker is
the trade name of a product supplied by GretagMacbeth. This information is given for the convenience of users of
this part of ISO 17321 and does not constitute an endorsement by ISO of the product named. Equivalent products
can be used if they can be shown to lead to the same results.)
NOTE 2 If measurements are not performed to 830 nm, there is a possibility that unwanted IR sensitivity will
not be identified during subsequent testing. One approach is to separately look at the transmittance of any UV and
IR blocking filters, the basic DSC filters, and the response of the detector itself. Another approach is illustrated by
the example in ISO 14524 which provides a method for evaluating the sensitivity of a DSC to IR radiation.
4.3.3 Test procedure
4.3.3.1 Test target illumination
4.3.3.1.1 For laboratory characterization using a reflection target
The spectral power distribution for illuminating the test target shall be photographic daylight D55,
as defined in ISO 7589. The illuminance at the target plane should be between 2 000 lx and 4 000 lx
and have a maximum variation of 1 % over the area being imaged. Annex A outlines a recommended
laboratory set-up for photographing a colour reflection test target.
The primary axis of the incident illumination should be approximately 45° to the normal to the centre
of the target area being imaged. Two or more illumination sources, which are equally spaced around the
normal to the centre of the area of the target that is being measured, should be used.
NOTE With the optical axis of the DSC normal to the test target, this will help to minimize the probability of
specular reflections entering the field of view of the DSC.
In qualifying the illumination source, particular attention should be paid to the rolloff in red response.
The spectral distribution index (SDI) described in ISO 7589 assumes a rolloff in red response which
is normal with silver halide films, but does not naturally occur with typical DSC sensors. If a DSC has
a long wavelength red response that is significantly different from that assumed in ISO 7589, the SDI
criterion is possibly not sufficient for qualifying the illumination source. ISO 14524:2009, Annex B, also
contains information about the relevancy of SDI calculations to the qualification of illumination sources
for DSCs. If there is some question about the relevancy of the SDI, the illumination source used should be
chosen so that its spectral power distribution matches that of the desired source as closely as possible,
in addition to meeting the SDI criterion.
If it is determined that an IR blocking filter is required for OECF determination, in accordance with
ISO 14524, the long wavelength red and infrared response of the DSC should be checked to determine if
the response is being appropriately dealt with by the DSC’s filters. If the DSC shows abnormally high long
wavelength red response (see the standard red rolloff in ISO 7589), or significant near infrared response,
additional filters can be used with the DSC at all times. It is also possible that the illumination source can
be emitting excessive amounts of infrared radiation, in which case the IR blocking filter should be placed
on the source, and the source requalified. If measurements of the DSC spectral response to wavelengths
from 840 nm to 1 100 nm were obtained, these spectral response values can be used to further qualify
the infrared rejection of the DSC. The ratio of the sum of spectral responses from 360 nm to 730 nm to
the sum of spectral responses from 740 nm to 1 100 nm should be greater than the ISO DSC luminance
[3]
dynamic range, as measured in accordance with ISO 15739 .
4.3.3.1.2 For in situ characterization
When the DSC is to be characterized in situ, the illumination source to be used for actual imaging shall
be used. This can be the digital still camera’s own illumination, studio illumination, backlighting for
transparent targets, or natural light (either artificial or daylight).
Where possible, the primary axis of the incident illumination for a reflection target should be
approximately 45° to the normal to the centre of the target area being imaged. Two or more illumination
sources, which are equally spaced around the normal to the centre of the area of the target that is being
measured, should be used.
NOTE Placing the optical axis of the DSC normal to the surface of a reflection test target will help to minimize
the probability of specular reflections entering the field of view of the DSC.
4.3.3.2 Camera focusing
The target distance is dictated by the size of the target, the field of view of the camera, and the focal
length(s) of the camera lens. These should be chosen such that the DSC is in sharp focus for the resulting
target distance.
4.3.3.3 Camera settings
For DSC characterizations performed in laboratory settings, the flash shall be turned off, any automatic
gain control shall be disabled, compression shall be minimized, and all user settings shall be recorded.
If possible, any digital white balancing should be turned off and any analog white balancing should be
fixed, so that variations in the analog white balance do not confound the white balance of the raw DSC
data revealed by the OECFs.
For DSC characterizations performed in situ, compression shall be minimized, and all user settings shall
be recorded. While local conditions can require the use of automatic gain control, if possible it should
be disabled. If possible, any digital white balancing should be turned off and any analog white balancing
should be fixed, so that variations in the analog white balance do not confound the white balance of the
raw DSC data revealed by the OECFs.
4.3.3.4 Image capture geometry
4.3.3.4.1 Full frame image capture
If full frame image capture is to be used, the test target should be framed within the DSC field of view so
that the set of fiducial marks appropriate for the aspect ratio of the DSC end up just inside the corners
of the captured image.
NOTE In the case of in situ characterization, the target can, of necessity, occupy a smaller portion of the
captured image.
6 © ISO 2012 – All rights reserved

4.3.3.4.2 Individual patch capture
If individual patch image capture is to be used, a black mask shall be used that has an aperture equal to
the size of an individual patch. The DSC or target shall be moved such that one patch at a time is seen by
the DSC at the centre of the image sensor area.
NOTE Individual patch capture can be used to minimize the effects of flare in the image capture step.
4.3.3.5 Data collection
Multiple images of the colour target, or individual patches, shall be recorded. These shall be analysed to
derive average DSC digital code values for each channel corresponding to each patch. In selecting sensor
element values to average, only those elements that are within the central 50 % of the area (70 % in
linear dimensions) of the image of each patch shall be included. At least, (64 × 64) pixels should be used
to analyse image data. Averaging shall be accomplished both within images and between at least three
DSC image capture samples.
4.3.3.6 Data reporting
Data reported shall include at least the following:
— make, model, serial number, etc., of target used;
— measured or reported spectral reflectance/transmittance factor of each patch;
— computed colorimetry of each individual test patch for photographic daylight D55 and any other
illumination used for testing;
— measured spectral power distribution of actual illumination source used;
— make, model, serial number, etc., of DSC(s) used for the test;
— mean and standard deviation of DSC code values corresponding to each patch;
— summary of DSC settings at time of image capture;
— DSC OECF used for analysis;
— capture geometry.
Annex A
(informative)
Recommended laboratory set-up for photographing a reflection
colour test target
A.1 General prerequisites
Three general prerequisites should be addressed in setting up a laboratory for photographing the
colour test target:
a) accurate positioning of the camera with respect to the test target;
b) proper illumination of the target;
c) capturing of essential data.
A.2 Camera positioning
A well-designed system for positioning cameras will accomplish the following objectives:
— optical axis of camera perpendicular to the target plane;
— optical axis of camera intersects centre of target;
— no skewing of target relative to image sensor;
— no keystoning of target relative to image sensor;
— easy to mount cameras in a repeatable fashion;
— easy to adjust shooting distance to fill camera’s field-of-view.
A.3 Target illumination
A well-designed system for illuminating the test target will accomplish the following objectives:
— simulated photographic daylight, or else spectral power distribution should be specified;
— geometry same as for measurement of spectral radiances reaching the camera;
— no specular reflections within the camera’s field-of-view;
— uniform illuminance at the target plane;
— direct illumination of the target, with negligible secondary contribution from scattered or
reflected light;
— regulated lamp voltage, with fixed level of direct current;
— easy way to confirm that the correlated colour temperature and illuminance at the target plane are
at their aim values.
8 © ISO 2012 – All rights reserved

A.4 Data acquisition
A well-designed system for acquiring data from images and for any associated metadata will accomplish
the following objectives:
— computer with adequate memory, DSC interfaces, and a visual display;
— means of downloading images from camera or from removable memory media;
— means of extracting statistical data from captured digital images;
— means of extracting metadata tags pertaining to camera make and/or model settings;
— means of recording all data into a tab-delimited text document;
— means of archiving all images with their associated measurements.
Annex B
(informative)
Digital still camera/sensitivity metamerism index (DSC/SMI)
B.1 General
In order to guarantee the colorimetric reproduction, a set of camera sensitivity curves is required to
[12]
obey the Luther condition , which requires a linear transformation of colour matching functions.
In practice, however, camera sensitivity curves deviate from the condition due to production reasons
of filters, sensor and optical elements. Such a camera will reproduce different sensor outputs for two
objects having the same tristimulus values but different spectral distributions.
DSC/SMIs are designed to give a measure for such potential colour error using the framework of
[6]
CIE Publication 13 . The indices consist of an average DSC/SMI and a special DSC/SMI.
Average DSC/SMI will give a measure of camera metamerism for ordinary reflective objects. In this
index, eight colour patches, defined in CIE Publication 13, represent reflective objects. Although a small
[4]
number of colour patches are used, simulations using ISO/TR 16066 verify that the average DSC/SMI
has a high correlation coefficient.
Special DSC/SMI is an optional measure by defining arbitrary objects depending on applications.
Although average DSC/SMI will give a statistically reasonable measure, the measure may not be reliable
for special objects such as highly saturated objects, fluorescent objects, and self-emitting colours.
Special DSC/SMI allows users to specify arbitrary objects in order to optimize for a specific application.
For instance, appropriate types of objects may be chosen from ISO/TR 16066.
Both measures give figures with the maximum of 100. The meaning of both indices is similar to
the one defined in CIE Publication 13. The colour mismatch indicated by a DSC/SMI value will be
comparable to the colour mismatch corresponding to the same colour rendering index value as defined
in CIE Publication 13. An index of 100 means that there is a very close match to the Luther condition. An
index of 50 corresponds approximately to the difference in colour rendering between D65 and a warm
white fluorescent (WWF) light source.
Both indices can be measured through either spectral basis (Method A) or test target basis (Method B)
as specified in Clause 4. Method B, however, can only be applied to a camera that can output linear
processing signals such as sensor raw data. On the other hand, Method B requires test targets having
appropriate spectral reflectances.
B.2 Measurement of DSC/SMIs using Method A
B.2.1 Step 1: Measurement of camera spectral sensitivities
Measure j channels of spectral sensitivities in accordance with 4.2.
[5]
IEC 61966-9 provides an alternate procedure that can be used where the requirements of 4.2 cannot
be fully met.
B.2.2 Step 2: Selection of light source
Assign an appropriate light source. Illuminant D55 should be used as default.
10 © ISO 2012 – All rights reserved

B.2.3 Step 3: Calculation of sensor outputs and tri-stimulus values
Calculate tristimulus values X , Y , Z , and sensor outputs O , O ,., O using the following equations.
i i i I,i 2,i j,i
Here, R (λ) is one of the spectral reflectances of colour patches, and s (λ) is the sensitivity of the jth
i j
channel ( j < 8). The spectral reflectances of the colour patches for average DSC/SMI and the spectral
distribution of the light source are listed in Table B.1.
Tristimulus values are calculated by the following equations:
XK= LR()λλ()x()λ (B.1)
ii∑
λ=380
YK= LR()λλ()y()λ (B.2)
ii
∑
λ=380
ZK= LR()λλ()z()λ (B.3)
ii∑
λ=380
where
K = (B.4)
Ly()λλ()
∑
λ=380
is the light source;
L()λ
are colour matching functions.
x()λ , y()λ , z()λ
Sensor outputs are calculated by the following equation:
OL= ()λλRs() ()λ (B.5)
ji, ij
∑
λ=380
Here, a spectral range from 380 nm to 780 nm with a 10 nm increment is suggested.
[6]
NOTE j is restricted to less than eight because only eight patches are used from CIE Publication 13 .
B.2.4 Step 4: Calculation of linearly optimized colour matrix
Calculate the tentatively optimized linear matrix using Equation (B.6):
−1
TT
AT= SSS (B.6)
()
where
XX X
 
18i
 
T = YY Y (B.7)
18i
 
ZZ Z
18i
 
OO O 
11,,11i ,8
 
S =   (B.8)
 
OO O
 jj,,18ij, 
 
A new set of estimated tristimulus values is calculated using:
TA= S (B.9)
kk
where
X
 
k
 
T = Y (B.10)
k k
 
Z
 k 
and
O 
1,k
 
S =  (B.11)
k
 
O
 jk, 
 
A measure for a system having more than three channels (e.g. complementary colour filter set such
as CMYG) may not be realized due to practical processing limitations. In such a case, a set of three
composite channels should be evaluated.
B.2.5 Step 5: Calculations of average DSC/SMI and special DSC/SMI
Calculate average DSC/SMI and special DSC/SMI as follows using the following equations:
Y
 3
ri,
*
L =116 −16 (B.12)
 
ri,
Y
 r 
1 1
 
 X 3 Y 3
 
* ri,,ri
a =500 − (B.13)
    
ri,
X Y
 r   r  
 
1 1
 
Y 3  Z 3
 
* ri,,ri
b =200 − (B.14)
    
ri,
Y Z
 r   r 
 
 
Y 3
* ki,
L =116 −16 (B.15)
 
ki,
Y
 k 
1 1
 
X Y
 3  3
ki,,ki
*
a =500 − (B.16)
    
ki,
X Y
 k   k  
 
12 © ISO 2012 – All rights reserved

1 1
 
Y Z
 3  3
ki,,ki
*
b =200 − (B.17)
    
ki,
Y Z
 k   k 
 
 
Colour difference ΔE is calculated by the following equation:
22 2
 2
* * * * * * *
ΔEL=−La+−ab+−b .
 
ab,ir(),ik,ir(),ik,ir(),ik,i
 
(B.18)
where
are the estimated tristimulus values of white (light source);
XY,,Z
kk k
XY,,Z
are the real tristimulus values of white (light source);
rr r
* * * are the estimated L*, a*, b* and tristimulus values of test target i;
La,,bXYZ,, ,
ki, ki, ki, ki,,ki ki,
* * *
are the real L*, a*, b* and tristimulus values of test target i.
La,,bXYZ,, ,
ri, ri, ri, ri,,ri ri,
Special DSC/SMI R and average DSC/SMI R are obtained by the following equations, respectively (note
i a
[6]
that the coefficient value of 5,5 is determined in order to align with CIE Publication 13 ).
RE=−100 55, Δ (B.19)
ii
RR= (B.20)
a ∑ i
i=1
B.2.6 Step 6: Non-linear optimization for Matrix A
Matrix A shall be optimized in a non-linear optimization technique in order to minimize the average
DSC/SMI with the initial matrix of the linear optimization. The resultant R , R shall be used.
i−NL a−NL
Note that Excel sheet “DSC/SMI.xls”, available from http://standards.iso.org/iso/17321 as part of the
ISO17321-1-Tools_ver101.zip, can be used for the non-linear optimization with the Solver command.
B.3 Measurement of DSC/SMIs with Method B
B.3.1 Step 1: Capture of colour patches
[6]
Capture the CIE Publication 13 eight patches (and also arbitrary colour patches if necessary) as
described in Clause 4.
NOTE A commonly available commercial source of such target patches that may be used for this application
is the Munsell colour book. The Federal Institute for Materials Research and Testing (BAM) of Germany also
offers reflection and transmissive test targets, accompanied by measurement data, meeting the requirements of
[6]
CIE Publication 13 . This information is given for the convenience of users of this part of ISO 17321 and does not
constitute an endorsement by ISO of these products.
B.3.2 Step 2: Calculation
Follow Step 3 to Step 5 as described in B.2. The data reported in 4.3.3.6 can be used for sensor output in
Equation (B.5).
B.4 Data reporting
Average DSC/SMI should be reported with a measurement method (Method A or B). If the illuminant
differs from D55, the light source should be identified. When Method A is used, spectral range and
spectral increment used in Equations (B.1) to (B.4) should be reported. When Method B is used, spectral
distributions of objects and light source should be measured and reported.
Special DSC/SMIs should be reported with spectral distributions of sample objects in addition to the
items required for average DSC/SMI.
Table B.1 — Spectral distribution of colour patches and D55 used for DSC/SMI
Spectral distribution
Wave-length
10BG
nm
7,5R 6/4 5Y 6/4 5GY 6/8 2,5G 6/6 5PB 6/8 2,5P 6/8 10P 6/8 D55
6/4
380 0,219 0 0,070 0 0,065 0 0,074 0 0,295 0 0,151 0 0,378 0 0,104 0 32,58
390 0,249 8 0,089 5 0,070 0 0,093 5 0,309 5 0,268 0 0,513 3 0,177 3 40,26
400 0,255 5 0,109 8 0,072 8 0,114 5 0,313 3 0,405 8 0,550 8 0,323 5 59,04
410 0,251 5 0,118 0 0,073 8 0,123 8 0,318 8 0,489 0 0,558 3 0,455 5 67,98
420 0,244 0 0,121 0 0,073 8 0,128 3 0,326 0 0,516 5 0,560 0 0,487 5 70,75
430 0,236 5 0,122 0 0,073 0 0,135 0 0,334 3 0,531 0 0,555 3 0,481 3 70,58
440 0,229 5 0,123 0 0,073 0 0,144 5 0,345 8 0,544 3 0,543 5 0,461 8 84,95
450 0,224 5 0,126 5 0,074 0 0,161 3 0,360 3 0,554 8 0,521 3 0,438 5 96,75
460 0,220 0 0,131 0 0,077 3 0,187 3 0,381 3 0,553 3 0,487 8 0,412 3 100,09
470 0,216 0 0,138 3 0,086 0 0,229 3 0,402 5 0,540 5 0,448 5 0,381 8 100,34
480 0,214 0 0,150 5 0,109 5 0,281 0 0,414 5 0,518 3 0,407 5 0,351 8 101,81
490 0,216 0 0,174 3 0,148 5 0,331 0 0,418 3 0,487 3 0,363 0 0,324 3 98,99
500 0,222 3 0,207 3 0,197 3 0,368 8 0,413 0 0,450 0 0,325 0 0,299 3 100,36
510 0,225 8 0,240 5 0,240 8 0,389 3 0,402 8 0,413 5 0,301 0 0,282 8 100,61
520 0,225 3 0,259 3 0,279 5 0,394 0 0,388 8 0,376 8 0,282 5 0,269 5 100,61
530 0,227 3 0,266 8 0,337 5 0,384 8 0,372 0 0,341 3 0,265 8 0,256 3 103,42
540 0,236 8 0,272 3 0,388 3 0,366 3 0,352 8 0,309 0 0,257 8 0,250 5 102,47
550 0,253 3 0,282 3 0,398 0 0,340 8 0,331 0 0,279 0 0,258 8 0,254 3 102,49
560 0,272 3 0,299 0 0,379 5 0,311 8 0,308 0 0,253 0 0,259 5 0,263 8 100,02
570 0,299 3 0,320 5 0,348 8 0,279 8 0,283 8 0,234 0 0,256 0 0,271 8 97,63
580 0,341 8 0,334 5 0,315 3 0,246 5 0,259 5 0,224 8 0,255 3 0,278 5 96,89
590 0,389 0 0,
...