ASTM E1455-97

NOTICE: This standard has either been superceded and replaced by a new version or discontinued.
Contact ASTM International (www.astm.org) for the latest information.
Designation: E 1455 – 97
Standard Practice for
Obtaining Colorimetric Data from a Visual Display Unit
Using Tristimulus Colorimeters
This standard is issued under the fixed designation E 1455; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (e) indicates an editorial change since the last revision or reapproval.
INTRODUCTION
This practice provides directions for correcting the results obtained with tristimulus colorimeters
when measuring the tristimulus values or chromaticity coordinates of colored displays. Tristimulus
colorimeters approximate the CIE color matching functions x¯ (l), y¯ (l), z¯ (l) to make these
measurements. The errors generated in measuring colors on a display may be minimized using this
practice.
1. Scope from Radiant Sources for Colorimetry
2.2 ISO/CIE Standard:
1.1 This practice is intended as an aid for improving the
CIE Standard Colorimetric Observers, ISO/CIE
accuracy of colorimetric measurements made with tristimulus
10527: 1991 (E) (International Organization for Stan-
colorimeters on visual display units, such as cathode ray tubes
dardization, Geneva, 1991)
(CRTs) and self-luminous flat-panel displays. It explains a
useful step in the analysis of colorimetric data that takes
3. Terminology
advantage of the fact that light from such displays consists of
3.1 Definitions: Unless otherwise stated, definitions of ap-
an additive mixture of three primary colored lights. However,
pearance terms in Terminology E 284 are applicable to this
it is not a complete specification of how such measurements
practice.
should be made.
3.2 Definitions of Terms Specific to This Standard:
1.2 This practice is limited to display devices and colori-
3.2.1 calibration, n— in reference to a tristimulus colorim-
metric instruments that meet linearity criteria as defined in the
eter, the process performed outside of this practice to adjust the
practice. It is not concerned with effects that might cause
tristimulus colorimeter to provide the best possible results for
measurement bias such as temporal or geometric differences
average or predefined conditions.
between the instrument being optimized and the instrument
3.2.2 optimization, n— in reference to a tristimulus colo-
used for reference.
rimeter, the process performed pursuant to this practice to
1.3 This standard does not purport to address all of the
adjust the tristimulus colorimeter or to interpret its readings to
safety concerns, if any, associated with its use. It is the
provide better results when applied to a particular display
responsibility of the user of this standard to establish appro-
device.
priate safety and health practices and determine the applica-
3.2.3 compatible, adj— in reference to a tristimulus colo-
bility of regulatory limitations prior to use.
rimeter, one so designed as to automate the procedure de-
2. Referenced Documents scribed in this practice.
2.1 ASTM Standards:
4. Summary of Practice
E 284 Terminology of Appearance
4.1 Tristimulus colorimeters comprised of three or four
E 1336 Test Method for Obtaining Colorimetric Data from
2 detector channels are, in general, not amenable to accurate
a Video Display Unit by Spectroradiometry
calibration that holds for all manner of usage with different
E 1341 Practice for Obtaining Spectroradiometric Data
illuminated devices and objects. This is because the spectral
This practice is under the jurisdiction of ASTM Committee E-12 on Appear-
ance and is the direct responsibility of Subcommittee E12.06 on Appearance of Currently available through the U.S. National Committee of the CIE, c/o Mr.
Displays. Thomas M. Lemons, TLA-Lighting Consultants, Inc., 7 Pond Street, Salem, MA
Current edition approved June 10, 1997. Published September 1997. Originally 01970-4819. Also included in ASTM Standards on Color and Appearance, Fifth
published as E 1455 – 92. Last previous edition E 1455 – 96a. Edition, 1996.
Annual Book of ASTM Standards, Vol 06.01.
Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.
NOTICE: This standard has either been superceded and replaced by a new version or discontinued.
Contact ASTM International (www.astm.org) for the latest information.
E1455–97
830 nm
responsivities of their detector channels do not exactly match
X 5 k F~l!x¯~l!dl (1)
*
360 nm
the defined Commission Internationale de L’Éclairage (CIE) x¯
830 nm
(l), y¯(l), z¯ (l) functions. Factory or subsequent calibration
Y 5 k F l y¯ l dl (1)
~ ! ~ !
*
360 nm
reflects judgments and compromises that may not be readily
830 nm
apparent. Nevertheless, this practice provides guidance on how
Z 5 k F~l!z¯~l!dl (1)
*
360 nm
such a tristimulus colorimeter may be optimized for use with a
particular video display device, providing better accuracy with
where:
that device than its more general calibration provides. An
k is 683 lm/W for emissive devices, such as displays, and x¯ (l),
optimization matrix transforms the instrumental (measured)
y¯ (l), z¯ (l) are color-matching functions. While the standard
CIE X, Y, Z values into adjusted X, Y, Z values that are closer
definition of X, Y, Z requires the use of the CIE 1931 2°
to the ideal. This matrix is determined by reference to a
color-matching functions, the mathematics described in this
colorimeter with higher intrinsic accuracy. The method derives practice would also be applicable to any other set of color-
from the fact that the color stimulus functions from display
matching functions, such as the CIE 1964 10° functions.
devices are linear combinations of three primary functions and
6.1.3 In practice, color measurement instruments compute
are not entirely arbitrary.
X, Y, Z by the summation of the signals as measured through
the various filters, each signal being multiplied by an appro-
5. Significance and Use
priate calibration factor. In matrix notation:
5.1 This practice may be applied when tristimulus colorim-
F
eters are used to measure the colors produced on self-luminous
X C C C . C F
m X1 X2 X3 Xf 2
video display devices such as CRTs and flat-panel displays,
Y C C C . C F
5 (2)
m Y1 Y2 Y3 Yf 3
F G F G
Z C C C . C A
m Z1 Z2 Z3 Zf 3 4
including electroluminescent (EL) panels, field emission dis-
F
f
plays (FEDs), and back-lit liquid crystal displays (LCDs). This
where:
practice is not meant to be a complete description of a
F , F , F , through F are the electrical signals from the f
procedure to measure the color coordinates of a display. Rather,
1 2 3 f
filtered detectors and the C are calibration coefficients. X ,
it provides a method for obtaining more accurate results when ij m
Y , Z have subscripts to indicate that they are measured
certain conditions are met. It may be used by any person
m m
values rather than ideal ones.
engaged in the measurement of color on display devices who
6.1.4 In this practice, we presume that the color measuring
has access to the requisite equipment.
instrument is linear: that each signal F is strictly proportional
5.2 This practice defines a class of tristimulus colorimeters a
to the received optical power, that any zero-offset (background
that may be said to be compatible with this practice.
in darkness) is removed, that the proportionality for signal F
a
6. Background of Practice
is not affected by the value of signal F , and in the case of
b
6.1 Colorimetry: closely packed detectors (such as charge-coupled device
6.1.1 Color measurement instruments consist, in general, of
(CCD) detector elements) no signal F spills over and affects
a
means to measure radiometric power as transmitted through a signal F as it approaches saturation. These presumptions are
b
number of bandpass filters. Most commonly, electrical devices amenable to experimental verification using methods beyond
are used to measure the filtered light. They may be used with
the scope of this practice (2).
different filters in succession, or multiple devices may be used
6.1.5 The values of the matrix elements C may be deter-
ij
concurrently. In instruments called spectroradiometers, the
mined using criteria that depends on the design and intended
radiometric power is measured through a large number (typi-
application of the instrument. The full extent of this subject is
cally 30 to 500) of narrowband filters. (Practice E 1341
beyond the scope of this practice. However, in general, for
describes how a monochromator or polychromator (spec-
spectroradiometers (f ’ 30 to 500), C reflects the tabulated
Xj
trograph) may be employed to filter and measure light in
value of x¯ (l) near the center wavelength of Filter j as well as
separate bands on the order of 1-nm wide.) In instruments
the spectral responsivity of the corresponding detector channel.
called tristimulus colorimeters, the radiometric power is mea-
(Likewise, C and C reflect y¯ (l) and z¯(l), respectively.) For
Yj Zj
sured through three or four wideband filters. These filters may
tristimulus colorimeters, the choice of C is discussed further,
ij
be constructed from dispersive elements (prisms and gratings)
below. As a general matter, the instrument designer should
or from materials with selective spectral transmission or
choose passbands and matrix elements that balance accuracy,
reflection. The latter may be either uniform or comprised of
sensitivity, and other design requirements.
different patches, in a mosaic pattern, that provide the desired
6.1.6 Tristimulus colorimeters are generally designed with
overall effect.
filters that are intended to match the spectral responsivities of
6.1.2 No matter how many filters are used, or in what
their detector channels to the CIE x¯ (l), y¯ (l), z¯ (l) functions.
manner, the goal of the measurement process is to determine
For such an instrument,
tristimulus values X, Y, Z, as defined by ISO in its Stan-
X C 00 F
m X1 1
dard 10527 and the CIE in its publication No. 15.2 (1). For
Y 0 C 0 F
m 5 Y2 2 (3)
F G F GF G
Z 00 C F
light with a color stimulus function F(l),
m Z3 3
where:
the non-zero C matrix elements represent adjustable gains of
Boldface numbers in parentheses refer to items in the list of References at the ij
end of this practice. the detector channels. However, the x¯ (l) function has two
NOTICE: This standard has either been superceded and replaced by a new version or discontinued.
Contact ASTM International (www.astm.org) for the latest information.
E1455–97
distinct lobes. This may be dealt with by splitting x¯ (l) into dimensional subspace is spanned by the F’s of all possible
x¯ (l) and x¯ (l), each with a separate filter (F and F , color stimulus functions following Eq 6. Further, the mapping
long 1 2
short
respectively). For such an instrument, of F into (X , Y , Z ) space by Eq 2 remains three dimen-
m m m
sional. In other words, there is a one-to-one mapping of the
F
X C C 00
m X1 X2
F vector (a, b, c) onto (X, Y, Z) by application of Eq 1; and, for
Y 00 C 0
m 5 Y3 (4)
F G F G
F
3 4 a particular instrument with a fixed calibration matrix C, there
Z 000 C
m Z4
F
is also a one-to-one mapping of the vector ( a, b, c) onto (X ,
m
Alternatively, the z¯ (l) function may serve the role of
Y , Z ). From this we deduce that a matrix R exists that can be
m m
x¯ (l) since they have a similar shape,
short used to translate (X , Y , Z ) values into actual (X, Y, Z)
m m m
values.
X C 0 C F
m X1 X3 1
Y 0 C 0 F
m 5 Y2 2 (5)
F G F GF G 6.3.2 A colorimeter that takes advantage of this fact must
Z 00 C F
m Z3 3
provide means for implementing the matrix R. That is, all f
In all of these cases, it is difficult to realize an exact match
filtered detector signals should contribute linearly toward the
between the CIE color-matching functions and the actual
computation of each output, X , Y , Z , instead of using
m m m
spectral responsivities of the corresponding detector channels.
different detectors for each output. This idea was reported as
This means that no choice of C will provide perfect calibra-
ij long ago as 1973 by Wagner (4), and it has been expanded
tion for all applications of the instrument. The criteria for
upon and rediscovered by others since then (5-9).
setting the C might not be well documented for a particular
ij
6.3.3 On the basis of this property, a tristimulus colorimeter
instrument.
can be optimized for use on a self-luminous display by the
6.1.7 It is generally believed that spectroradiometers, with
proper derivation of a matrix R for that display. We proceed on
their many detector channels, may be calibrated to yield
the assumptions that the components are sufficiently stable, and
superior measurements of X, Y, Z for diverse applications.
that similarly built displays have similar enough spectral
Nevertheless, the relative simplicity of tristimulus colorimeters
primaries to make a derivation of R worthwhile. However,
and their commensurately lower cost have made them popular
these assumptions should be quantified before accuracy claims
where the highest accuracy is not required.
are made in any specific situation.
6.2 Self-Luminous Displays:
6.3.4 On the basis of this property, a tristimulus colorimeter
6.2.1 A self-luminous display, such as a CRT, an electrolu-
designed for use with displays need not produce signals F that
minescent (EL) panel, a field emission display (FED), or a
are close to CIE tristimulus values. Signal/noise may be
back-lit liquid crystal display (LCD) generates colored light by
improved by matching the spectral responsivities of the filtered
the proportional superposition (addition) of primary colored
detectors to the emission spectra of the primary colors. In such
lights F (l), F (l), F (l). The subscripts represent red,
r g b
designs, it is especially important to use a matrix R that is
green, and blue, the primary colors of an additive set. An
specific to the particular F (l), F (l), F (l).
r g b
arbitrarily colored patch on the visual display has one and only
one color stimulus function F(l),
7. Optimization
F~l! 5 aF ~l! 1 bF ~l! 1 cF ~l! (6)
r g b
7.1 General:
where a, b, c are coefficients that are determined by the
7.1.1 Given the existence of a matrix R, how is it deter-
display electronics.
mined? Experimentally, the problem is one of comparing the
6.2.2 The display electronics vary a, b, c over the face of the
data X, Y, Z from a reference colorimeter with the data X , Y ,
m m
display in order to generate a colored image. For this practice,
Z from the colorimeter being optimized, for a number of color
m
we presume that the display electronics may be set to make a,
samples at different display setting
...