UNDERSTANDING

Datacolor

GLOSS
COMPENSATION

A Technical Report On Gloss and Datacolor's Gloss Compensation

What Is Gloss?

When light strikes an object, most of the light enters the object but a small amount is
reflected off the surface of the object. This surface reflected light, which is responsible for the
shiny or bright appearance of an object, is the appearance attribute called gloss.

Figure .1

Figure 1 shows a beam of light of intensity I incident upon an optically smooth film. A
small percentage of this light is reflected off the surface but most of the light enters the film.
The reflected beam S leaves the surface at an angle rl which is equal to the incident angle i.

The light beam R that enters the film is changed in

direction or refracted upon entering the different material. This
occurs because there is a difference in refractive index' between
the two mediums. In the example in Figure 1, the incident
beam I is in air with a refractive index of 1.00 and the object
material has a refractive index of 1.50. The amount of bending
of the light beam depends on the refractive indexes of the
mediums and is related to the sine of the angles by Snell's Law.

1
See Appendix A for Explanation of Reflective Index.

1
Specular Reflection

Specular reflection refers to the mirror-

like reflection of light from a smooth surface.
When light strikes a perfect mirror surface, all of
the incident light is reflected off the surface.
The angle of the reflected light beam is equal to
the angle of the incident light beam.

Figure 2 - Specular Reflection

For objects that have optically flat surfaces but are not mirrors or metals, the
specular reflectance differs in magnitude but not in direction. The amount of specular
(mirror-like) reflection from a smooth surface will depend on the refractive index of the
material and the angle that the light strikes the surface. The angle of the reflected light
beam will still be equal to the incident light beam. A theoretical value for the specular
reflection can be calculated using the Fresnel2equation, which will give the specular
value for any refractive index or angle of incidence. If all objects were optically smooth
or flat, the measurement of specular reflection would be straightforward.

An optically smooth or flat surface is glossy in appearance. Because of the

directional nature of the specular reflection, the observer can position the sample to
eliminate the specular reflection from his field of view. This permits the observer to
only see the light from the interior of the sample. The observer in this case sees the
color with maximum saturation because the surface reflected light is not included in the
view. The specular reflectance has the same chromaticity or color as the light source
and since it never penetrated the surface, it dilutes the object's saturation.

An example of this observation would be a newly waxed black car. The car
looks very bright and shiny but also appears very black and highly saturated. The
brightness is the result of seeing the specular reflection and the saturated black is
observed when we change viewing angles and eliminate the specular. We are then
observing the color without any surface reflected light in our field of view.

2
· See Appendix B for Explanation of Fresnel Equation.

2
Diffuse Reflection
Most real objects do not have perfect
mirror-like surfaces. As the surface
becomes rougher, the specular reflection
becomes diffuse. This means that the
light is reflected back at many different
angles, instead of only the mirror angle.
This rough type of surface is referred to
as a matte surface. The specular
reflection is not reflected back at the
same angle as the incident angle. Instead
the reflected light is spread out over a
range of angles depending on the makeup
of the surface.
Figure 3 - Diffuse Reflectance

The viewer of a matte object cannot eliminate the surface reflection by changing the
viewing angle. Because of the diffuse nature of the surface reflectance, the observer
will always see some of this surface reflectance. This diffuse surface reflected light,
which has the chromaticity or color of the light source, is added to the light from the
interior of the object and tends to lighten and desaturate the color of the object.

An example of this observation is to take the newly waxed black car in the previous
example and sand blast or abrade the surface of the car. Instead of seeing a shiny and
deeply saturated black color, we would now see a duller and dark gray color. The
pigmented paint layer did not change. The paint surface is now a matte finish and the
specular reflection is diffuse. When we observe the car, we will always have some of
the specular in our field of view, no matter what our angle of view. The car now looks
lighter and less saturated than when it was newly waxed.

Figure 4 - Glossy Surface Figure 5 - Matte Surface

Viewer A does not see the specular while Viewer B does Both Viewer A and B see the specular reflectance
see it. It is dependent on the angle of view. because of the diffuse nature of the reflectance.

3
Specular Gloss

Specular gloss is a measurement of the brightness of the reflected image. A

numerical value is determined based on the ratio of the specular reflection from a
sample, to the specular reflection of a standard material, under the same geometric
conditions. This numerical value is the specular gloss value and it is usually obtained
from an instrument called a glossmeter.

The glossmeter consists of a light

source that directs the fight to the sample at a
fixed angle. A light detector is placed at the
mirror angle to measure the intensity of the
reflected image. Figure 6 shows the physical
setup of a gloss meter. The 60' angle is the
most commonly used angle with the 20' used
for high gloss materials and the 85' used for
low gloss or matte materials. The gloss angle
should always be specified with any reading.
(Note. AS7M D-523 recommends 20 o for 60
o
gloss > 70 and 85o for 60o gloss < 10.)
Figure 6 - Secular Glowmter

The measured reflectance is not used directly, but is compared to the light
reflected from a black glass standard. The light reflected from the sample is divided by
the light reflected from the standard black glass. This value multiplied by 100 is defined
as specular gloss.3

It is important to note that the specular glossmeter is measuring only specular

gloss. There are other types of gloss. Two objects may have the same specular gloss
but appear distinctly different to an observer. An example of this situation can be seen
where one sample shows a very clear and sharp reflected image and another sample
shows a blurred and hazy image. The samples may have the same specular gloss but a
visual appraisal will indicate a difference.

The above is an example of distinctness-of-image gloss. Other types of gloss

include contrast gloss, absence-of-bloom gloss, sheen, etc. The important point to
realize is that gloss, especially as it relates to appearance, includes other aspects in
addition to simple specular gloss as measured by a glossmeter.1
3
See Appendix C for more information on Specular Gloss.
4
The Gloss Problem

The lack of agreement between visual and instrumental evaluations of color samples
that have different gloss has been a major problem in making color appearance
judgments. A human observer can easily change the viewing conditions and the
viewing angles in order to make an appearance judgment, taking "into account both
color and gloss simultaneously.

Color measuring instruments are not as versatile as the eye, since measurements are
made at fixed illumination and viewing conditions that do not correspond to the way we
see objects. The human eye has much greater resolving power than most instruments
and can distinguish very small appearance differences between objects.

The basic problem encountered when viewing samples having different gloss can be
illustrated in the following example. Suppose we have two black samples; one with a
high gloss surface and the other with a low gloss or matte surface. If these samples are
measured with an integrating sphere instrument with the specular included (type SCI
d/0), the color difference will show that there is a small color difference between the
matte sample and the high gloss sample. A visual assessment of these samples will
indicate that there is a larger color difference. The high gloss sample will appear much
darker and more saturated than the low gloss sample.

In order to understand this problem, one needs to become aware of the

relationship between instrument geometry and surface reflectance. We have already
seen the visual effects of objects with different surface characteristics. We will now
examine the effect of instrument geometry on samples with differing gloss.
5
 Instrument Geometry

Instrument geometry refers to the physical makeup of the color measuring

instrument; relating specifically to the nature and position of the light source, and the
angles of the viewing optics. The instrument geometry will greatly affect the
measurement data and the color difference calculations derived from this data. There
isn't one instrument geometry that can duplicate the large number of viewing conditions
used in day to day color evaluations worldwide. There are, however, some standardized
instruments geometries that have been defined by the CIE2 and have been used
successfully for years in both color matching and quality control applications.

Standard Illuminating and Viewing Conditions3

6
Specular Component Included (SCI) - d/0

This condition uses an integrating sphere to diffusely illuminate the sample and
uses a viewing angle near the normal to the sample surface (i.e. perpendicular to the
surface). The most common viewing angle used is 8. The 8o angle approximates the 01
angle very closely since the specular reflection for 8o and 0o is almost the same
according to the Fresnel equation.

The measurement data will include any specular reflection that is reflected into
the viewing optics at the viewing angle. For a glossy sample the light from the area of
the sphere at the mirror angle to the viewing optics (-8 degrees for d/8), will enter the
optics. For a matte sample, the diffused surface light, which can be from any area of the
sphere, will enter the optics.

Figure 7 - Gloss Sample SCI Figure 8 - Matte Sample SCI

Figure 7 and 8 show what happens when a glossy and matte surface is measured
with this geometry. These two objects are identical in color and differ only in surface
gloss. In Figure 7, the 4 arrows on the left of the sphere depict the incident light. In
reality, light is coming from all areas of the sphere wall. Since this is a glossy object,
Figure 7 also shows the light reflected from the sample by the 4 arrows on the right of
the sphere. Again this is a simplification to help explain what is happening. In reality,
the light is being reflected back at many angles. The glossy object reflects the specular
directly into the viewing optics. This is shown by the bold arrow in Figure 7.

In Figure 8, we are again only showing 4 incident light beams. Since this is a
matte object, the incident beams are diffused and are not reflected in the mirror direction
as in Figure 7. The specular reflectance does not enter the optics directly but because of
the matte surface, the diffused specular reflectance indirectly enters the viewing optics.
Since the specular is included in each of the measurements, the measurement of the
glossy object will be the same as the measurement of the matte object.

7
Specular Component Excluded (SCE) - d/0

This geometry adds the use of a specular port or gloss trap to the sphere. The
specular port is an opening in the sphere wall placed at the mirror angle to the viewing
optics. For a d/8 design, the specular port would be placed at -8 degrees from the
normal relative to the viewing angle. When this port is open, no light from that area is
available to reflect off the sample at the mirror angle and enter the optics.

Figure 9 - Gloss Sample SCE Figure 10 - Matte Sample SCE

Figure 9 shows an example with a glossy sample. There isn't any light available at
the mirror angle (because of the open port) to enter the optics. Notice that the bold
arrow is not there. The measurement includes only the diffuse reflection from the
interior of the sample. No specular reflection enters the viewing optics.

Figure 10 shows an example with a matte sample. In this case the diffused surface
light, which can be from any area of the sphere, will enter the optics. Notice that even
though the bold arrow is not there, the diffuse reflection from the other arrows (incident
light) will enter the optics. The measurement includes both the diffuse surface reflected
light and the diffuse reflection from the interior of the sample. Some specular reflection
enters the optics.

The glossy sample and the matte sample will have different measurements using this
geometry. A color difference assessment will indicate that the glossy sample is darker
and more saturated than the matte sample. The use of this geometry will improve visual
agreement, especially in reference to the direction of the color difference, but the
magnitude or size of the color difference typically does not always agree with a visual
assessment. SCE has not been found to be a preferable geometry over SCI, except for
specialized needs, in any major industrial application.

8
0/45 and 45/0 Geometries

0/45 instruments illuminate the sample at the normal to the surface (i.e. 90 deg to
surface), with the viewing optics at an angle of 45 degrees to the normal. Some 0/45
designs also offer multiple viewing optics (circumferential) that place the viewing optics
at fixed positions around the sample. 45/0 instruments illuminate the sample at 45
degrees with the viewing optics near the normal. Some 45/0 designs place the light
sources in a ring around the sample at 45 degrees with the viewing optics near normal.
The CS3 45/0 is of this type.

The 0/45 and 45/0 geometries exclude the specular component of the sample
reflectance completely for a smooth or glossy sample but pick up some of the diffuse
specular reflection of a rough or matte sample.

Figure 11 - 4510 Gloss Surface Figure 12 - 45/0 Matte Surface

Figure 11 shows a 4510 configuration for a glossy sample. The illuminating light is
incident on the sample at 45 degrees and the specular reflection is reflected back at 45
degrees. The specular reflection does not enter the viewing optics. Figure 12 shows a
45/0 configuration for a matte sample. For a rough surface or matte sample, the diffuse
specular reflection from the surface is reflected in all directions and some will enter the
optics.

Because observers usually rotate samples to avoid the specular, the 45/0 and 0/45
geometries correspond more closely to the viewing conditions that are normally used to
view and compare colors. It is generally accepted that these geometries provide better
agreement with visual assessments for samples with substantial gloss differences.
Instruments with a 45/0 or 0/45 geometry have been used in quality control applications
but have found limited utility in computer formulation and correction applications.
9
Why isn’t 0/45 the Preferred Geometry?

Even though the 0145 and 4510 geometries may provide the best agreement with
visual assessments in cases where there is a significant gloss difference, they are not the
preferred geometries for general work; especially when using a color computer for
formulation and correction. The main reason for this is that the 0/45 or 45/0 design will
always picks up any gloss variations, however slight, in its measurements. These gloss
differences will be included in the measurement data without any user intervention
pertaining to the validity of the data.

This gloss sensitive characteristic of the 45/0 design can cause problems in
building colorant data files for a color matching system. If the primary samples used to
make the colorant data file have gloss differences due to pigment loading, resin types, or
even fingerprints or other surface irregularities, the primary data will be linked to these
arbitrary gloss levels. They will not represent the true characteristic of the primary
colorants.

Another possible problem area is in making batch corrections. All gloss

differences are included in the spectral measurement data and will be interpreted as
color differences. If the batch gloss varies from batch to batch, a correction program
will interpret these gloss differences as color differences between the standard and the
batch, and will attempt to make adds to compensate. This could lead to incorrect or
occilating adds. The SCI geometry masks these differences so that only the actual color
differences are acted on.

The 45/0 and the 0/45 geometries, although more representative of the way we
view real objects, are generally not best suited for overall work in the computer color
field. There are some specialized areas such as quality control where these geometries
have been used successfully. Unless there are specific requirements, it is probably better
to use a SCI geometry, especially for new users of color systems.

10
What Is the Datacolor Gloss Function?

The Datacolor Gloss Function combines the virtues of both the d/0 and 0/45 or 45/0
geometries. It allows us to take a reflectance curve obtained from a spectrophotometer
of d/8 SCI geometry and adjust the curve based on the sample's gloss value to be
roughly what it would read on the other geometry. This adjustment moves the
reflectance curve from the dimension or space of the specular included instrument,
which we call “measurement space", to a new dimension which we call “visual space”.
The "Visual space" curve will be a better representation of the color and will agree more
closely to the way we actually see the sample.

The function is derived from data obtained from measurements of a series of black
samples, consisting of different gloss levels from matte to high gloss. A d/8 SCI and
0145 instrument were used to gather the necessary data. The d/8 SCI instrument will
measure basically the same reflectance values for all these samples while the 0/45
instrument will produce varied reflectances based on the gloss of the samples. A visual
appraisal of the samples indicated much greater visual color differences between the
lower gloss samples (O to 30 gloss) than the higher gloss samples (30+ gloss). This
indicated that the visual change that is observed when viewing the gloss series was not a
linear relationship. In other words, equal changes in gloss value did not produce equal
visual changes.

The differences between the SCI data and the 0/45 data at each gloss level, when
plotted, showed a non-linear relationship that seemed very similar to the perceived
visual difference. It followed that it should be possible to adjust a SCI curve based on
this relationship and obtain a new curve that would be similar to a curve actually
measured on a 0/45 instrument. Since 0/45 measurements have usually been in better
agreement to visual perception, this new curve should also agree better to visual than the
SCI curve.

This method was tried and found to agree much closer with visual appraisals than
SCI. In fact, the results were close to the data obtained from the 0/45 instrument. The
benefit of this approach is that we can apply the gloss compensation only when we need
it. We retain all the benefits of the SCI sphere geometry but we have the ability to
correct certain curves based on this "visual" function. The function requires input of
either a gloss value obtained from a glossmeter or a gloss factor obtained from a
CS5/CS3 instrument.
4
See Appendix D for a graphical explanation.

11
Color Difference Before Gloss Correction

The color differences shown above are from specular included measurements of a high
gloss black sample and a low gloss matte sample. The gloss value of the high gloss sample is
91.0 @ 60' and the gloss value of the matte sample is 9.9 @ 60'. Visually the high gloss sample
looks much darker than the matte sample.

Color Difference After Gloss Correction

After the gloss correction is applied the color differences agree with the visual appraisal.
The matte sample is 5.69 units lighter than the glossy sample. The hue difference remains
essentially the same. If this sample pair were measured on a 0/45 instrument, we would get
almost the same result.
12
 CS5 and CS3 Gloss Determination

The Datacolor Gloss Factor

The CS5 and CS3 with automatic specular ports can be used to determine a gloss factor
of a sample. This gloss factor is an indication of the gloss of the sample, and can be interpreted
like a gloss value from a specular glossmeter. The gloss factor is a relative number, which is
correlated to actual specular gloss values of a set of gloss standards. The gloss factor will not
always agree with values obtained from a glossmeter. The agreement between a gloss factor
and an actual glossmeter reading will be dependent on the physical characteristics (refractive
index and type of surface) of the calibrating gloss standards and the material being measured.5

Each instrument is calibrated for gloss at Datacolor by reading a common set of gloss
standards. The actual glossmeter values of these gloss standards are entered into the calibration
program. The set of gloss standards is then measured with the instrument and the correlation of
gloss values is established. Even though the gloss factor may not always agree exactly with a
glossmeter gloss value, the factor is close enough for use in the gloss compensation option.

If a closer correlation to actual glossmeter values is desired, the instrument can be gloss
calibrated on a gloss ladder made of the material that will be commonly measured. For
example, a paint company could make up a black paint gloss series and calibrate on that. The
returned gloss factors would then correlate much closer to the actual glossmeter values of
paints.

The CS5 and CS3 gloss factors can be used successfully as the input to Datacolor gloss
compensation programs. They can also be used to quantify the gloss of a product and to
determine gloss agreement. It is important to note however, that if product specifications
specify that gloss be measured according to ASTM D 523 or that a specific glossmeter
geometry be used, you NLUST use a glossmeter. The gloss factor is not a replacement for a
glossmeter when a glossmeter value is specified. In the absence of specific gloss specifications,
the factor can be a convenient way to determine gloss agreements and the factor itself can be
defined into a specification between users of CS5's and CS3's.

5
See Appendix E for Graphical Explanation.

13
 APPENDIX A

Refraction of Light

Consider a ray of light

passing from medium A into
medium B. If the speed of light is
greater in A than in B, the ray is
bent toward the normal NN to the
surface of separation, as shown in
Fig. 13. If it is less in A than B,
the ray is bent away from the
normal.
Figure 13

A constant n of the two media is defined by the ratio:

speed of light in A sin i

n = ----------------------------- = ---------
speed of light in B sin r

where i is called the angle of incidence, r the angle of refraction,

and n the index of refraction of B relative to A. When A is a
vacuum, n is called the (absolute) index of refraction of medium B.
Since the speeds of light in air and vacuum are practically equal,
the refractive index of a substance relative to air is practically equal
to that relative to a vacuum. 4

14
 APPENDIX B

With equation #2, the specular reflectance of an

optically smooth film can be calculated. All that is
required is the angle of the incident light and the refractive
index of the film.

Example: Incident Angle = 8 degrees

Refractive Index = 1.5

Specular Reflectance = 4.00062%

15
 APPENDIX C

SPECULAR GLOSS MEASUREMENT

R (sample)
G --------------------- x 100
R (std glass tile)

where R (sample) = %R of Sample Measured

R (std glass tile) = Specular reflectance of black glass
tile of known refractive index
G = Specular Gloss Value

According to ASTM D 523, highly polished, plane, black

glass with a refractive index of 1.567 should be assigned a specular
gloss value of 100 for each geometry. The gloss value for glass of
any other refractive index can be computed from the Fresnel
equation. For example, glass with a refractive index of 1.527
would have gloss values of 89.2, 93.6, and 99.4 for the 20, 60, and
85-degree geometries.5

When calibrating a specular glossmeter, you will measure a

black glass tile and adjust the meter to a predetermined value. This
value is supplied with the black glass standard and is based on the
refractive index of the black glass. If the black glass had a
refractive index of 1.527, you would adjust the meter to read 93.' 6
for the 60-degree geometry. If the black glass had a refractive
index of 1.567, you would adjust the meter to read 1 00 for the 60
degree geometry.

16
 APPENDIX D

Y Values of Black Gloss Ladder

Measured SCI, SCE, 0/45

A series of black samples varying in

gloss from matte to high gloss were
measured with an SCI, SCE, and 0/45
geometry. The Y values are plotted in
Figure 14. The SCI values, Curve A in
Figure 14, form a horizontal line across the
chart indicating that there isn't any change
between the samples. The SCE values,
Curve B in Figure 14, indicate a change
but a rather linear change. The 0/45
values, Curve C in Figure 14, indicate a
change but a very non-linear change.

When viewing these samples arranged in order of increasing gloss, you will notice a
much greater color difference (mainly lightness/darkness) between adjacent samples at the
low gloss end than at the high gloss end. The visual appraisal of these samples agrees
more closely with Curve C, the 0/45 relationship, than with Curve B, the SCE relationship,
or Curve A, the SCI relationship.

Delta Y for SCI-0/45 and SCI-SCE

Figure 19 shows the delta Y values

versus gloss. The delta Y's are the
differences between the specular included
measurements of the black gloss ladder
and the 0/45 and specular excluded
measurements of the black gloss ladder.
The delta Y function for SCI-0/45 is used
to adjust an included curve to visual space.
The gloss value is chosen from the x aids
and a vertical line is drawn until it
intersects the SCI-0/45 curve. From this
point a horizontal line is drawn to the y-
axis and the delta y is read. This is the
amount to subtract from the SCI curve.
17
 APPENDIX E

CS5/CS3 Gloss Determination and Datacolor Gloss Function

Figure 16 - Gloss Determination and Gloss Function

Figure 20 shows how a gloss factor is obtained from an instrument and how that gloss factor is
applied to obtain a corrected curve. The instrument takes an SCI and SCE read of the sample to obtain
a delta Y value. This delta Y value is then used to obtain the gloss factor. In the top graph in Figure
20, this is shown graphically.

The bottom graph of Figure 20 shows the delta Y (SCI-0/45) versus gloss graph inverted 180'.
This is done to show how the returned gloss factor (or glossmeter value) is used to obtain the delta Y
used to adjust the SCI curve. From the gloss factor or value, a line is connected to the SCI-0/45 curve.
From this point a line is drawn to the y axis to obtain the amount of Y or actually %R to subtract from
the included curve.

18
 References

1. Richard Hunter and Richard Harold, 7he Measurement of Appearance, (New York,
John
Wiley & Sons, 1987), p. 78.

2. CIE (Commission Internationale de I'Eclairage), International Commission for

Illumination.

3. Deane B. Judd and Gunter Wyszecld, Color in Business, Science and Industry, Third
Edition, (New York, John Wiley & Sons, 1975), P. 123.

4. Daniel Schaum, Yheory and Problenu of College Physics, (New York, McGraw-
Hill, 1961), p. 219.

5. ASTM, AS7M Standards on Color and Appearance Measurement, (Philadelphia,

PA, ASTM, 2nd edt 1987), D 523-85 p. 13.

19