September 10, 2009 / Vol. 7, No. 9 / CHINESE OPTICS LETTERS 869
Color appearance and visual measurements for
color samples with glos s effect
Jian Ma (
êêê
èèè
)
1,2
, Haisong Xu (
MMM
°°°
ttt
)
1∗
, M. Ronnier Luo
2
, and Guihua Cui
2
1
State Key Laboratory of Modern Optical Instrumentation, Zhejiang University, Hangzhou 310027, China
2
Department of Color Science, Leeds University, Leeds LS2 9JT, UK
∗
E-mail: chsxu@zju.edu.cn
Received December 15, 2008
We assess the color appearance of the samples with different inks on glossy substrates, five kinds of paper
with different gloss levels. The color samples are measured using spectrophotometers under different
illuminating/viewing geometries and v isually estimated using the psychophysical method of magnitude
estimation. The results of the two approaches are compared through the color appearance mo del of
CIECAM02. The experimental data analysis indicates that the 0/45 and 15/0 geometries can be u sed
to describe the two major aspects of gloss effect, the enlargement of color gamut, and the reduction of
lightness. The agreement for hue attribute between instrumental measurement and visual assessment is
better than those for colorfulness and lightness.
OCIS codes: 330.4595, 330.1730.
doi: 10.3788/COL20090709.0869.
It is well known that the gloss of printed material has
a significant effect not only on its appearance but also
on the perceived surface color
[1]
. Matte surface tends
to have lower chroma and higher lightness compared to
those with gloss effect, when the samples are measured
without the specular light
[2]
. While with the presence of
sp e cular light, the results tend to be opposite. To as sess
samples with gloss effect, both measurements with and
without specular reflection should be evaluated to give
full compr e hens ions of the gloss factor. In this letter, the
color samples with gloss effect are visually assessed and
physically measured under different illuminating/viewing
geometries. Through the CIECAM02 color appea rance
model, the results of different geometries against visual
data are evaluated and compare d
[3,4]
.
Five kinds of paper with different gloss scales were
selected to be substrates, w hich were Niceday recycled
plain paper (P), Epson photo quality paper (PQ), Epson
premium glossy photo paper (PG), Olmec glossy inkjet
paper (OG), and Epson ultra glossy photo paper (UG).
All specular gloss values of the five substrates were mea-
sured using a Sheen Tri-micro Gloss 160/T gloss meter.
As listed in Table 1, the substrates were measured under
the geometry of 60
◦
first and those whose gloss values
were under 10 were measured again under the geometry
of 85
◦
[5]
.
Both pigment and dye inks were inspected, which were
provided by Clariant Co., Ltd. There were five differ-
ent colors: the CMYK (cyan, magenta, yellow, and key)
primaries and a color made by mixing equal amounts of
CMY together, all of which had two different concentra-
tions, resulting in 10 colors for pigment and dye inks,
respectively. All the 20 ink samples for the two types of
Table 1. Gloss Values of Selected Substrates
Geometry P PQ PG OG UG
60
◦
3.5 2.3 37 29 62
85
◦
5.5 2.9 – – –
inks were applied on to five different substrates using
a K-bar with spiral of 0.1 µm. In total, 100 color sam-
ples were pre pared, with the size of 3 × 3 (inch).
The visual assessment was performed using a view-
ing cabinet in a dark room, under the light source of a
D65 simulator, by the psychophysical method of magni-
tude estimation. A panel of eight observers with normal
color visions, including four males and four females, par-
ticipated in the visual assessment, seven of which had
academic backgrounds in color science and the other one
was well trained befo re participation. The observers were
asked to make quantized estimations of the test samples,
which were displayed randomly, for their appearance at-
tributes with maximum gloss effect by comparison with
the given references . The references for lightness and col-
orfulness assessment corresponded to N10 from Munsell
color or der system and 3040-R20B from NCS (Natu-
ral Color System) color order system, respectively. All
observers were clearly instructed about their tasks and
fully adapted to the illumination of the cabinet. 20 of
the 100 samples were assessed twice for the insp e c tion of
intra- variation of the observers. In total, each observer
made 120 a ssessments, all of which were divided into two
sessions to avoid the fatigue of the observers.
All samples were then measured using three different
types of spectrophotometers, Greta g Macbeth CE7000A,
X-Rite 939 (0/45), and Dataco lor FX1 0. Specula in-
cluded (SPIN) and s pecula excluded (SPEX) g e ometries
for Gretag Macbeth CE7000A, 0/45 for X-Rite 939, a nd
10 kinds of geometries for Datacolor FX10 were adopted.
The intra- and inter-variation were analyzed firstly, to
check the reliability of the collected data. The perfor-
mance of the observers was reported in terms of STRESS
(standardized residua l sum of squares) values
[6]
, which
can be calculated using
STRESS =
P
(E
i
− F
1
V
i
)
2
P
F
2
1
V
2
i
!
1/2
× 100, (1)
1671-7694/2009/090869-04
c
2009 Chinese Optics Letters