Lighting Guide 11:

Surface reflectance and colour

Its specification and measurement for designers

The Society of
Light and Lighting
NPL I
National PhysicalLaboratory
Lighting Guide 11:
Surface reflectance and colour
Its specification and measurement for designers

The Society of
Light and Lighting
222 Balham High Road, London SW12 9BS, UK
TeL020 8675 5211. Fax: 020 8673 3302. e-mail: [email protected]. www. cibse.org
The Societyof Light and Lighting is part of the
Chartered Institution of Building Services Engineers

NPLI
National Physica' Laboratory
National Physical Laboratory
Queens Road, Teddington, Mlddx TW1 1 OLW
Tel: 020 8977 3222. Fax: 020 8943 6458. www.npI.co.uk
This document is based on the best knowledge available at the time of publication.
However, no responsibility ofanykind for anyinjury, death, loss, damage or delayhow-
ever caused resulting from the use of these recommendations can be accepted by the
Chartered Institution ofBuilding Services Engineers, the Society ofLight and Lighting,
the National Physical Laboratory, the authors or others involved in its publication. In
adopting these recommendations for use each adopter by doing so agrees to accept full
responsibility for any personal injury, death, loss, damage or delay arising out of or in
connection with their use by or on behalf of such adopter irrespective of the cause or
reasonthereforeand agreesto defend, indemnify andhold harmless the CharteredInsti-
tution of Building Services Engineers, the Society of Light and Lighting, the National
Physical Laboratory, the authors and others involved in their publication from any and
all liability arising out ofor in connection with suchuse as aforesaid and irrespective of
any negligence on the part of those indemnified.
Note from the publisher
This publicationis primarily intendedto give guidance. It is not intended to be exhaus-
tive or definitive, and it will be necessary forusers ofthe guidance givento exercisetheir
own professional judgementwhen deciding whetherto abide by or depart from it.

The rights ofpublication or translation are reserved.

No part ofthis publication niay be reproduced, storedin a retrieval system or transmitted

in any form or by any means withoutthe prior permission ofthe publisher.

© 2001 The Society of Light and Lighting and the National Physical Laboratory
AllNPL material usedin this Guide is © Crown Copyright 2001 andhas been reproduced
by kind permission ofHMSO.

The Society is part ofCTBSE which is a registered charity, number278104.

ISBN 1 903287 14 6

Typeset by the Society of Light andLighting.

Printed in England by Page Bros, Norwich.

Foreword
This publication is one of a series produced by the Society and before it the CIBSE
Lighting Divisionon different aspects of lighting. This particular guide deals with the
specification and measurement of surface reflectance and colour. It will be of value to
designers and specifiers, including architects and the designers ofinteriors andproducts
as well as lighting specialists. It is seen as a work ofreference as well as providing a tool
for practical assessment ofsurfacereflectance.
The taskgroupresponsible for the publication, whose namesare listedbelow, includes
both designers and scientists thus giving the work an important value. The Society
gratefully acknowledges theirtime and expertise and in particular the inputprovided by
the staff ofthe National Physical Laboratory.
The Task Group is grateful to the CIBSE Publications Manager and the Society
Secretary for their help in the final stages ofthe project.

DavidLoe
Task Group Chairman

Task Group
David Loe, Chairman(Building Research Establishment)
Teresa Goodman (National Physical Laboratory)
lain Maclean (lighting consultant)
Keith Miller (DPA LightingConsultants)
Julie Taylor (National Physical Laboratory)
Lighting Guide: Surface reflectanceand colour 5

Contents

1 Introduction 6

2 Light reflecting properties of surfaces 6

3 Light reflecting properties of building materials 9

3.1 Reflectance of building materials 9
3.2 Colour of building materials 10
3.3 Surface finish of building materials 10
3.4 Surface deterioration and maintenance 11
3.5 Depreciation of luminaire performance 12

4 Surface colour specification 12

4.1 Munsell system 12
4.2 Natural Colour System 15

4.3 DIN system

16
4.4 BS 5252: 1976: Framework for colourco-ordination
for buildingpurposes 17
4.5 RAL design System 20

5 Measurement of reflectance and colour 20

5.1 Measurement using a spectrophotometer 20
5.2 Measurement using a colorimeter 21
5.3 Measurement using luminanceand Illuminance meters 22
5.4 Field measurements using the reflectancesample card 22
5.5 Visual assessment 23

6 Reflectance and colour standards 23

7 Surface appearance 24

8 Explanation of terms 29

References 31

Reflectance sample card description 32

Reflectance samplecard Inside back cover

Lighting Guide. Surface reflectanceand colour 7

1 Introduction
All lit environments, whether daylit or electrically lit, will contain an element of re-
flectedlightwhich will contribute to individual illuminance and luminance valueswithin
a space. The amount and nature ofthe light reflected from a surface will depend on the
reflectance propertiesof the material, which include its total reflectance, its reflectance
factor, the colourof the surface and the surface finish. These surface properties, in addi-
tion to the nature of the light illuminating it, will determine, in a physical sense, its
appearance including its luminance, luminance factorand chromaticity. But the appear-
ance ofan elementwithin a scene will also, to some extent,be determinedbythe lit environ-
mentofwhich it is part. This is throughbrightness and colouradaptation and constancy,
as well as the contrast, both brightness and colour, between adjacent surface elements.
The guidewill explainthe various properties ofreflecting surfaces and how they can
be assessed and measured. It will also provide guidance on the reflectance of typical
building materials and a range of coloursamples. These can be used to assess, by com-
parison, the reflectance value of materials.
It is expected that the guidewill providea valuable reference source as well as day to
day information for all designers, but particularly lighting designers, architects and inte-
rior designers.

2 Light-reflecting properties
When lightis incidenton a surface, some may be absorbed, some transmittedand some
reflected. The reflected light may be contained within a very small range of angles —
termed specular reflection — or spreadover a rangeofangles termeddiffuse reflection.
For a perfectly specular surface, such as a mirror or a polished metal surface, all the
incident light is reflected in the same plane as the incident light, and at an equal and
opposite angle to the surface normal as the incident light (see Figure 1). There is no
diffuse reflectionand no losses. For a perfect diffuser (white blotting paper and matt
emulsionpaint are good approximations), the light is reflected so that the surface ap-
pears equally bright in all directions and there is no specular reflection (see Figure 2).
Incidentlight Reflected light Intensityof the
reflected light
varies as cos e
regardless of
angleof
Incidentlight incidence

Specular surface Diffusing surface

Figure 1. Reflection from a perfectly specular Figure 2. Reflection from a perfectly diffuse
surface. The reflected beam is at the same surface. The luminance of the surface is
angle to the surface as the incidentbeam and independent of the angle of view, i.e. the
is also of the same intensity luminance is a constant and the intensity of the
reflected lightat an angle 9 to the
perpendicular to the surface varies as cos 9.
8 Societyof Light and Lighting and National Physical Laboratory

Reflected light Figure 3. Most types ofsurface exhibita

peaks around combination of diffuse and specularreflection.
the specular
reflectionangle The proportionof specularreflection
determines how 'glossy' or 'shiny' the surface
Incident light appears.

Glossy'surface

For most surfaces the reflection is neitherperfectly specular or perfectly diffuse — these
surfaces may be termed 'glossy' (significant specular reflection), 'semi-matt' (little
specular reflection) etc. (see Figure 3).
The reflectance properties of a material are determined by its structure. Specularly
reflecting materials have a smooth, homogeneous surface whichintroduces no distortion
into thepath ofthe reflected light. Diffusely reflecting materials have a 'rough', inhomo-
geneous structure which scatters the reflected light in all directions. In most cases light
will also penetrate belowthe surface and, for anynon-homogeneous material, scattering
will occur. Ifthe scattering particles exhibit selective absorption in the visible region, a
colour is imparted to the reflected light and the material appears coloured. Consider the
case of a coloured gloss paint layer, for example, in which coloured pigments are sus-
pended in a clear gloss lacquer. Specular reflections occur from the surface ofthe lac-
quer but a reasonably high proportion of the light penetrates to the particle layer below,
giving rise to coloured,diffuse, reflections.The overall effectis a coloured,high gloss finish.
Ifwe plot the way in which the proportion of light reflected varies with wavelength
for a white or grey surface, a flat line is obtained. For a coloured surface on the other
hand, troughs (corresponding to regions of low reflection) and peaks (regions of high
reflection) are seen. A 'red' surface, for example, generally reflects red lightbut absorbs
blue and greenand its reflectance profile might be similar to that shown in Figure 13 of
Section 6. However it is not just the reflectance properties of a surface that affects its
colour appearance. The spectral composition ofthe light illuminating the surface is also
important. If green light falls on the red surface shownin Figure 13 of Section 6 then it
appearsblack- there is no red lightpresentin the incident lightto be reflected. Daylight
and the light from tungsten filament lamps contain radiation at all visible wavelengths
and therefore allowus to see all shades ofcolour. Other lamps, suchas fluorescent lamps,
emit most oftheir light at a relatively restricted range ofwavelengths and can therefore
'distort' the colour appearance ofa surface. Some fluorescent lamps, for example, have
very little red radiation and make red surfaces look 'subdued'. The general colourren-
dering indexR,, ofa lamp is a measure ofhow well colours can be seen or distinguished
from one another. Tungsten lamps have a general colourrendering indexof 100; fluores-
centlampscan have an indexof-50-95, accordingto theirindividualspectral distribution.
Thereare manyotherfactorsapart from its reflectance and colourandthe lightsource
with which it is illuminated which will affect the visual appearance of a surface. These
include its texture, the level and uniformity ofthe illumination and the colourand lumi-
nance of adjacent surfaces. These issues are discussed further in Section 7.
Lighting Guide: Surface reflectanceand colour 9

3 Lighting properties of building materials

One ofthe purposes ofthis publication is to help designers to assess the lighting proper-
ties ofbuilding surfaces. The surfaces mightbe interior or exterior, although for exterior
surfaces this will normally only be required for appearance purposes as with floodlight-
ing. Where an exterior surface is used to reflect lightfor illumination purposes then the
designer is advised to consider it in a similar way to an internalsurface.
In assessing a building surface there are threemain things to consider, i.e. the surface
reflectance, its colour and its surface finish. A further consideration is how the material
will change through its life. For example it will get dirty, it may change colourand even
the surface finish may change. Thesechangeswill affect the lighting properties and will
alsoneed to be taken into accountat the lighting design stage.

3.1 Reflectance of building materials

Building surfacescome in a wide range of finishes from the diffuse matt surface of a
plasteredwall, through textured surfaces like brickor stone as well as carpets, and on to
specular surfaces like glass and metal. From an indirect illumination aspect it is nor-
mally the diffuse reflectance that is used. However the designer should be aware that the
reflection factorand the appearance ofthe surface can change depending on how it is lit.
For example if a textured surface is lit obliquely then the textured appearance will be
enhanced but if it is lit from all angles then the texture will be suppressed.
The previous section has described diffuse reflectance, and how this is measuredis
described in Section 5. A further way of assessing diffuse reflectance is to use the Re-
flectance Sample card provided, the use of which is also described in Section 5. When
assessing diffuse reflectance the sample will need to be lit with diffuse light andprefer-
ably with the illurninant to be used in the finaldesign. Ifa surface is all onemateriallike
a concreteor rendered wall then the diffuse reflectance of only this material needsto be
determined. If, however, the surface is a combination of two materials then both will
need to be taken into account. For example a brick wall is a combination of the bricks
and the mortarthat holds them togetherand typically the mortar can be as much as 14%
ofthe total area ofthe wall. In this case, and any othercombination ofmaterials, the area
weighted reflectance will need to be used. This is where the diffuse reflectance of the
individual materials is weighted by the particular areas for a diffuse reflectance of the
whole surface to be determined,
e.g. Area weighted reflectance ofa brick wall.
Area ofbrick = A1
Area of mortar = A Iii

Reflectance of brick = R1,

Reflectance of mortar = R

Area weighted reflectance ofbrick wall = (A1, x R1,) + (A,,, x R,,,)

A1 +Ain

This approach must be used wherever there is a combination of surface materials or

features whichhave differentreflectances and it is the combined diffuse reflectance that
is required. The brick wall has already been mentioned but it applies to other surfaces
like a perforated panel which is a combination of the main surface and the very low
10 Societyof Light and Lighting and National Physical Laboratory

reflectance of the holes. It will also apply to whole room surfaces like a wall that in-
cludes the wall surface, perhaps a door of different colour and windows — all the ele-
ments will need to be taken into account.
With regard to the actual diffuse reflectance of building materials, they cover a very
wide rangeand the designer is advisedto seek out as accurate a value as is possible. Also
for major surfaces actual samples should be assessed preferablyunder the illuminant to
be used. For example bricks, depending on the particulartype, could vaiy anywhere
between0.2 and0.7 while timberis likely to fall between 0.1 and 0.4. Clearglasson the
otherhand is typically around 0.15. Textured surfaces like carpets will generally have a
relatively low reflectance unless they are very light in colour. Paintedsurfaces will also
vary widely. For example a white emulsion paintedplastersurface could have a reflect-
ance as high as 0.8 while a dark painted surface could be as low as 0.1. The diffuse
reflectance of some materials, such as paints, can oftenbe obtainedfrom the manufac-
turer. Howeverifmanufacturers datais not available then oneofthe methods described
in Section 5 should be used.

3.2 Colour of building materials

Building materialscan come in a wide range of colours as well as many different
reflectances. Bricks for example range from dark reds through to pale yellows as well as
the dark blue of engineering bricks. Stone also comes in many colours and shades but
like brick it depends on the particular natural material. Man-made materials, however,
depend on the materials used in its manufacture or on pigments introduced to create a
particularcolour. Cement for example can come in many colours depending on the de-
sired effect required by the architect. Forpainted surfaces then the manufacturer's speci-
ficationwill usually include the colourspecification as well as the diffuse reflectance. It
is important that an accurate system ofcolourspecification is used to ensurethe desired
effect as determined by the architect or interior designer — see Section 4.

3.3 Surface finish of building materials

The surfacefinish ofbuilding materials vary from the smoothmatt surface ofplaster to
the highly textured surfaces of shuttered concrete. There are also highly polished sur-
faces which include some glasses, polished stone like marble andmetal panels finished
with a powder coating. The nature of the surface finish will affect how the light is re-
flected from it, but this will also be affected by the waythe surface is lit. As has already
been explained, if a surface is matt then it will scatter light evenly in all directions. If
however, the surface is textured andthe surface is lit obliquely then only the elements of
the surface which receive light will reflect light, whereas those that do not will be in
shadowand appear 'dark'. This combination oflight and shade will be effective in en-
hancing the appearance of the surface texture but it will reduce the average luminance
which can be interpreted as a reductionin its average reflectance. Similarly, if the sur-
face is specular then it may be possible to see a reflectedimage ofthe light source. But
this will depend on the geometry between the surface, the illuminant and the viewing
position. As has been mentioned earlierthe designer will need to decide on the primary
purpose ofthe material. If it is for effect, then the lighting should complement the sur-
face material to enhance the effect and the light reflecting properties as it affects the
general illumination ofthe room can be dealt withby consideration ofthe diffuse reflect-
ance ofthe material. Ifhoweverthe direction and intensity ofthe lightreflected from the
Lighting Guide: Surface reflectanceand colour 11

095 i I I I I

09

085 Clean
L)

c0 08
U
o
a)
075

07

065 Dirty

06

055
6 12 18 24 30 36
Elapsedtime(moraths)

Figure 4. Reflectancereduction against elapsedtime, from CIE Publication 97.

surface is important, then an appropriate reflection factor or factorswill be required. As
with the other lighting properties of materials much will be achieved by examining,
experimenting and measuring actualmaterial samples to determine the lighting effects.

3.4 Surface deteriorationand maintenance

Throughout the life ofa building the surfaces will deteriorate throughtheiraccumulation
ofdirt whichin turn will reduce the amount ofreflected light.How muchthiswill be and
at what rate is difficultto determine withoutlong termtests whichin most cases will not
be possible. Nonetheless the designer must consider this process ifthe lighting levels are
notto reduce to a pointthat is unacceptable. The magnitude ofthese changes is governed
by the extent ofdirt deposition which will depend on the nature ofthe surface and the
cleanliness ofthe atmosphere. It will also dependon the importance ofthe surface with
regardto inter-reflected lighton illuminances withinthe space. Afurther element in this
is the regularity that the surface is cleanedor refurbished in some otherway. As a guide
to the likelychange in diffuse reflectance whenthe surface is used primarily as a source
ofinter-reflected lightthe graphshown in Figure 4 hasbeenprovided. It shows the likely
reflectance reduction factorfor clean, average and dirty environments against the length
oftime betweencleaning periods.The data for this have been taken from the CIE Pub-
lication 97, which doesnot take into account the effect ofthe surface texture to retain or
accumulate dirt. The overall message must be that the designer at least considers the
likelyeffect ofdirt accumulation inthe particularsituation andthe effect onilluminances
within the space. The building owners, or their clients, should also be encouraged to
clean the room surfaces as regularlyas possible.

3.5 Depreciation of luminaire performance

Theaccumulationof dust and dirt on all room surfaces is fully covered abovebut it must
not be forgotten that the same dirt deposition may also effect the performance of lamps
and luminaires. Detailed guidance on the practical calculation ofthe effect ofdirt on and
in luminaires and the resultantloss in performance may be made by reference to section
4.5.2 in the 1994 edition ofthe CIBSE CodeforInterior Lighting.
12 Society of Light and Lighting and National Physical Laboratory

4 Surface colour specification

In terms of light reflection from building surfaces for illumination purposes, it is the
reflectance ofthe surface that is important. Howeverthis takes no account of its colour.
Lightreflectedby a colouredsurface will be tinged with that colour, but except in strictly
controlled environments, where accurate colour representation is required, this is rarely
a problem. The colour appearance ofa surface will be affected by the spectral radiation
distribution of the light source illuminating it, which can be described simply by its
chromaticity co-ordinates, either on a CIE 1931 x andy diagram or on a CIE 1976 u' and
v' uniform chromaticity diagram, see Figures 5 and 6. Chromaticity co-ordinates can
alsobe used for surfacecolours but only when illuminated by a particular lightsource. In
this case the co-ordinates combine the spectral reflectance distribution ofthe surface and
the spectral radiation properties of the light source. Chromaticity co-ordinates are also
the basis by which a light source's colour rendering quality can be determined — for
furtherdetails see the 1994 edition ofthe CIBSE Codeforinterior lighting.
Paintmanufactures often use names to describe the colourof a producthut thesecan
cause confusion. For example can a user be really sure of the exact colour of a paint
describes as 'emperor' or even 'daffodilyellow'. To overcome this a numberofnumeri-
cal systems have been devised. The following are the most commonly used.

4.1 Munsell system

The Munsell systemis the oldest of those currently employed. This was devised by
A.H.Munsell originally in 1905 andhas been extendedandrefined in various ways since
(Nickerson, 1976). The system provides a three dimensional matrix and is shown in
Figures 7 and 8. The surface colours are defined in terms of threeperceptual attributes:
hue, value and chroma.
Hue describes the apparently dominant part ofthe spectrum occupied by the colour,
e.g. red, as distinct from yellow or blue. The various hues are locatedaround the periph-
ery ofa circle located on a horizontal plane andpassing throughthe Munsell solid. There
are five principal hues and five intermediate hues:
Principal hues: red (R), yellow (Y), green (G), blue (B), purple (P)
Intermediatehues: yellow red (YR), green yellow (GY), blue green (BG), purple
blue (PB), red purple (RP).
Chromais the strengthof the colour and increases radiallyfrom the centre ofthe solid
which is the point ofneutralgrey (zero chroma) to a maximum whichdependsupon the
hue and reflectance of the surface. Surfaces havingzero chroma, and therefore no hue,
are denoted neutral (N).
Value is plotted verticallythrough the colour solid and measures the lightness of the
surface from 0 (perfect black) to 10 (perfect white). Value is another way of describing
reflectance, but differs numerically from reflectance in an important respect: each ofthe
three Munsell scales is divided in such a waythat equalintervals ofhue, value or chroma
denote approximately equal steps in perceived contrast (thoughthe value scale doesnot
have the same spacingas the hue or chroma scales).
The approximate relationship between reflectance R (%) and Munsell value V is given
by the equation:
R = V(V— 1)
Ughng Guide Surface reflectanceand colour 13

0,8
CLL
Sf0
HROMflWY
0,7
DGRAM

48

800

48
30005

__/j/ /
'1
04 0::oo57' of4

\
000OK
0,3 4Sf N 43

— 5M4

42 40

P
450
5 Tq.o'I
A a. gk ,d.,a..AapaI25$5
Pba..AIdayAghIwlth*aa...*.*A
'I oooAI.n,p.UlS.. ofUSAK
450

0 0,3 0,4 0,5 0,0 0,7 0,8

x
Figure 5. CIE (x,y) chromaticity diagram 1931

06

05

04

03

02

01

0 01 02 03 04 05 06 07
U,

Figure 6 CIE uniform (u',v) chromaticity diagram 1976.

14 Societyof Light and Lighting and National Physical Laboratory

Rh te

50
1ORe

-cc;
scale
5YR
cl:'{ 5Y

Black

Figure 7. Schematic diagram of the Munse// co/oursystem (reproduced with kind permission of
the Munse// Co/or company Ba/timore, USA).

Figure 8. Munse// co/our so/idwith s/ice removed, showing boundary of 5Y (ye//ow)hue p/ane
(reproduced with kind permission of the Munse/l Co/or company,Baltimore, USA).

TheMunsell co-ordinatesfor a coloured surface comprise hue, value andchroma, in that

order. For example, Munsell reference 2.5GY 6/8 indicates that the hue is 2.5GY, a
distinctly yellowish green, the valueis 6, which meansa reflectance ofabout0.3, andthe
chroma is 8 (moderately saturated but not startling).
Figure 8 shows the bounding surface of the Munsell colour solid. This indicates the
relative positions ofthe highest possible chromas at each hue andvalue. Thushigh chroma
yellows have high value (reflectance) while high chroma reds and blues have low
reflectances. (Web site: http://www.rnunsell.com)
Lighting Guide: Surface reflectance and colour 1 5

Figure 9. Natural Colour System hue circle

(reproducedby kind permission of the
Scandinavian Colour Institute)

4.2 Natural Colour System

The Natural Colour System (NCS) was developedin Sweden by Hard and Sivik (1981).
In this system, colours are scaled according to their degree of resemblance to the six
elementary colours: white, black, yellow, red, blue and green. Of these, the four chro-
matic colours are those in which no trace of any other can be seen, e.g. the red that
contains no trace of either yellow or blue and so on. These four are placed 900 apart on
the hue circle and form the basis for red-green andyellow-blue opponent axes, see Fig-
ure 9. The white and black are the pure colours of this type that contain no trace ofthe
otheror ofany ofthe chromatic elementary colours. Any givencolourcan have elements
of no more than two of the chromatic colours plus white and black. For example, an
orange hue mightbe saidto bear 30% ofred and 70% ofyellow; the sumofthe elements
ofthe chromatic elementary colours is always 100. The hue ofthis orangesample would
then beY3OR. The degree ofresemblance to black is calledblackness (s), that to white is
called whiteness (w), that to yellow is called yellowness (y), and so on. The degree of
resemblance to a maximum, or completely chromatic colour, is calledchromaticness (c).
By dividing an elementary scale into 100 steps, it is possible to get a measure of the
degree ofresemblance to the elementary colours.
The Natural ColourSystem (NCS) triangle(Figure 10) is therefore a diagram inwhich
the resemblance ofa colourto white,blackandthe maximum chromatic colouris plotted
in terms ofwhiteness, blackness and chromaticness, and
S + W + C = 100

Figure 10 illustrates the arrangement of samples in the NCS Atlas in a constant-hue

trianglewith cornerpoints w, c and s. Arraysofconstant chromaticness (c) are parallelto
the vertical axis. Constant blackness (s) and constant whiteness (w) arrays lie on lines
parallel to the sloping sides of the triangle and do not pass through the corners unless
blackness or whiteness is zero. It is clear that it is only necessary to specifythe hue, s and
c in orderto specifythecolourcompletely. However, neitherthe hue circlenorthe con-
stant-hue trianglesshow spacings that are closeto visual equality. This reflectsthe very
16 Societyof Light and Lighting and National Physical Laboratory

differentguidingprinciples uponwhich it was based compared withthe Munsellsystem.

The NCS atlas contains 1412 colour samples and more recently, ICI (Paints) has intro-
duced an atlas based on the NCS system. (Web site: http://www.ncscolor.com)

Figure 11. Colour solid of the DIN system T = hue, S = saturation, D = darkness (reproducedby
kind permission of Ellis Horwood).

4.3 DIN system

TheDIN (DeutschesInstitut für Normung) systemwas developedin Germany by Rich-
ter and Witt (1986). The three variables usedare hue (7), saturation (5) and darkness (D),
quoted in that order andthe arrangement is shown in Figure 11. The systemuses a stand-
ard daylight designated D( light source, and CIE tristimulus values for the samples
available. Hunt describes tristimulus values in 'Measuring Colour' (1991). There are 24
principal hues, having values of T= 1 for a yellow, via reds,purples, blues, and greens to
a yellow-green of T= 24. These principal hues were chosento represent equalhue differ-
ences between adjacent pairs all round the hue circle.
The second variable, saturation, is a function ofthe distance from the pointrepresent-
ing the reference white on a chromaticity diagram, and for colours of the same reflect-
ance (not the same darkness) equal saturation represents equal perceptual differences
from the grey ofthe same reflectance.
The third variable is darkness, rather than lightness, andthis is not related in a simple
way to reflectance.
The coloursolid associated with the DIN system is formed by havingthe grey scale as
a vertical axis, with white (D = 0) at the top and black (D = 10) at the bottom. Different
hues are represented by 24 different planeswith one edge coincident with the grey scale
axis and the same angle (1 5°) between adjacent planes.
An atlas for the DIN system is available, knownas the DIN ColourChart: the sample
size is 20 x 28 mm. In the atlas, for eachhue page, colours of equal DIN saturation are
arranged in columns parallel to the grey axis, instead ofin lines radiating from the black
point.
Lighting Guide: Surface reflectanceand colour 17

4.4 BS 5252: 1976: Framework for colourco-ordination for building

purposes
A system widelyused in the building industryin the UK is the British Standards system
BS5252(1976) whichis a list of237 surface colours. Though cross-referenced to Munsell
co-ordinates, BS 5252 adopts another colournotation system specially developed to avoid
some ofthe anomalies ofthe Munsell system.
The three dimensions of colourin BS 5252 are: hue, designated by an even number;
greyness, designated by a letter, andweight, designated by an additional number, usually
odd.
12 distinct hues are identified plus 'neutral':
00 = neutral
02 = red-purple
04 = red
06 = warm orange
08 = cool orange
10 = yellow
12 = green-yellow
14 = green
16 = blue-green
18 =blue
20 = purple-blue
22 = violet
24 = purple
The estimatedgrey content ofsurface colours is represented on a five-point alphabetical
scale:
A = grey
B = near grey
C distincthue
D = nearly clear
B clear, vivid colour
Generally, greyness correlates well with Munsell chroma but the two concepts are not
interchangeable, for black (00E53) and white (00E55), having no greyness content, are
clear (group E) not grey (group A).
Surfaces having the same weight generally look equally light, but the correlation be-
tween weightand reflectance, even withina given greyness group, is uneven. In groups
A and B the correlation is excellent, in group C it is good, but in D and E it is poor. This
is because saturated redsand blues tend to look lighterthan yellows or greys ofthe same
reflectance; the phenomenon is knownas the Helmholtz—Kohlrausch effect (Hunt, 1991).
Individual coloursare identified by a combination ofa hue number, a greyness group
letterand aweightnumber, in that order. For example, 12B27 meansthat the huenumber
is 12 (greenish yellow). The greyness group is B (close to grey) but with a slight hue.
The weightnumber is 27, indicating a darkishtone. The colouris a dark olivegreen.
Subsets ofcolours suitable forparticularapplications have been picked from BS 5252
and published as separate British Standards see BS 4800, 4900, 490], 4903, 4904. For
example, the BS 4800 paint coloursarelisted in Table 1, with their approximate Natural
Colour System and Munsell equivalents and their reflectances.
18 Societyof Light and Lighting and National Physical Laboratory

Table 1. Approximate NCS and Munsell referencesand reflectances for colours in BS

4800: 1981
Greyness Colour Hue Approximate Approximate Approximate
group designation NCS Munsell reflectance
-- reference reference
A UUAUI eutra1- 1501-YO3R N 8.5 1)68
00A05 Neutral 3101-Y26R N7 0.45
00A09 Neutral 5301-R64B N5 0.24
00 A 13 Neutral 7501-R97B N3 0.11
10A03 Yellow 2002-YO3R 5Y8/0.5 0.60
10A07 Yellow 4302-YO9R 5Y6/0.5 0.33
10A 11 Yellow 6702-G98Y 5Y4/0.5 0.14
B 04 B 15 Red 0906-Y78R 1OR 9/1 0.79
04 B 17 Red 1409-Y83R 1OR 8/2 0.62
04 B 21 Red 38l0-Y76R 1OR 6/2 0.33
08 B 15 Yellow-red 0606-Y4IR 1OYR9.25/1 0.86
08 B 17 Yellow-red 1607-Y4IR 8.75YR8/2 0.64
08B21 Yellow-red 4107-Y41R 8.75YR6/2 0.32
08 B 25 Yellow-red 6308-Y4OR 8.75YR4/2 0.16
08 B 29 Yellow-red 8105-Y53R 8.75YR2/2 0.07
10 B 15 Yellow 0504-Y21R 5Y9.25/1 0.87
lOB 17 Yellow 181l-YO1R 5Y8/2 0.61
10B21 Yellow 4011-G99Y 5Y6/2 0.33
10 B 25 Yellow 621 1-G9OY 5Y4/2 0.16
10 B 29 Yellow 8305-G89Y 5Y2/2 0.07
12 B 15 Green-yellow 0807-G73Y 5GY 9/1 0.81
12 B 17 Gren-ycllow 1812-G75Y 2.5GY 8/2 0.61
12 B 21 Green-yellow 3915-G65Y 2.5GY 6/2 0.33
12 B 25 Green-yellow 6313-G57Y 2.5GY 4/2 0.15
12 B 29 Green-yellow 8207-G53Y 2.5GY 2/2 0.07
18 B 17 Blue 1704-B78G SB 8/1 0.62
18 B 21 Blue 4004-B57G 5B 6/1 0.34
18 B 25 Blue 6405-B 14G SB 4/1 0.16
18 B 29 Blue 8205-BO6G 7.5B 2/1 0.06
22 B 15 Violet 1000-N IOPB 9/1 0.81
22 B 17 Violet 1804-R58B 1OPB 8/2 0.60
C 02 C 33 Red-purple 1118-RO7B 7.5RP 8/4 0.62
02 C 37 Red-purple 3531-R17B 7.5RP5/6 0.23
02 C 39 Red-purple 5331-R21B 7.5RP3/6 0.10
02 C 40 Red-purple 7315-R24B 7.5RP2/4 0.07
04 C 33 Red 1019-Y86R 7.5R 8/4 0.62
04 C 37 Red 3632-Y85R 7.5R 5/6 0.23
04 C 39 Red 5136-Y87R 7.5R 3/6 0.10
06 C 33 Yellow-red 1517-Y35R 7.5YR 8/4 0.62
06 C 37 Yellow-red 4034-Y45R 5YR5/6 0.23
06 C 39 Yellow-red 6525-Y4OR 7.SYR 3/6 0.11
08 C 31 Yellow-red 0809-Y32R 1OYR9/2 0.81
08 C 35 Yellow-red 2430-Y24R 1OYR 7/6 0.46
08 C 37 Yellow-red 4340-YI8R 1OYR 5/6 0.23
08 C 39 Yellow-red 6724-Y22R 1OYR 3.6 0.10
lOC3l Yellow 0811-YI6R 5Y9/2 0.81
10 C 33 Yellow I 122-YO3R 5Y 8.5/4 0.71
10 C 35 Yellow 2536-G99Y 5Y7/6 0.45
10 C 39 Yellow 6921-G9SY 5Y 3/4 0.10
12 C 33 Green-yellow 1623-G72Y 2.5GY 8/4 0.62
12 C 39 Green-yellow 6626-G49Y 2.5GY3/4 0.10
Lighting Guide: Surface reflectanceand colour 19

Table 1 continued
Greyness Colour Flue Approximate Approximate Approximate
group designation NCS Munsell reflectance
reference reference
14 C 31 Green 0609-G12Y 5G 9/1 0.81
14 C 35 Green 2610-GO6Y 5G 7/2 0.45
14 C 39 Green 6520-G 5G 3/4 0.10
14 C 40 Green 8007-GOSY 5G 2/2 0.07
16 C 33 Blue-green 1613-B68G 7.5BG 8/2 0.60
16 C 37 Blue-green 4326-B57G 7.5BG 5/4 0.22
18 C 31 Blue 0704-B97G SB 9.25/1 0.84
18 C 35 Blue 2516-BO5G 7.5B7/3 0.42
18 C 39 Blue 6126-BO8G 7.5B3/4 0.10
20 C 33 Purple-blue II 17-R83B 5PB 8/4 0.63
20 C 37 Purple-blue 3827-R87B 5PB 5/6 0.23
20 C 40 Purple-blue 7415-R82B 5PB 2/4 0.07
22 C 37 Violet 3928-R6OB 10PB 5/6 0.22
24 C 33 Purple 1514-R3SB 7.5P 8/3 0.60
24 C 39 Purple 5431-R49B 7.5P 3/6 0.10
D 04D44 Red 2858-Y88R 7.5R4/l0 0.16
04 D 45 Red 3657-Y93R 7.5R 3/10 0.10
06 D 43 Yellow-red 2560-Y27R 7.5YR 6/10 0.33
06 D45 Yellow-red 4644-Y47R 5YR4/8 0,16
10 D 43 Yellow 2163-G97Y 5Y7/10 0,45
10 D 45 Yellow 3952-098Y 5Y 5/8 0.24
12D 43 Green-yellow 2954-G64Y 2.5GY 6/8 0.32
12D 45 Green-yellow 5043-G54Y 2.5GY 4/6 0.15
16 D45 Blue-green 5536-B51G 7.5BG 3/6 0.10
18 D 43 Blue 3536-BO9G 7.5B 5/6 0.22
20 D 45 Purple-blue 4938-R88B 5PB 3/8 0.10
22 D 45 Violet 4542-R63B IOPB 3/8 0.10
E 04E49 Red 0314-Y9IR 7.5R9/3 0.80
04 E 51 Red 0963-Y8IR 7.5R 6/12 0.33
04 E 53 Red 1777-Y8IR 7.5R4.5/16 0.18
06 E 50 Yellow-red 0742-Y32R 7.5YR 8/8 0.60
06 E 51 Yellow-red 0860-Y5OR 2.5YR7/11 0.46
06 E 56 Yellow-red 2962-Y43R 5YR 5/12 0.24
08 E 51 Yellow-red 1178-YI6R 1OYR7.5/l2 0.51
10 E 49 Yellow 0823-G87Y IOY 9/4 0.79
10 F. 50 Yellow 0848-Y SY 8.5/8 0.64
10 E 53 Yellow 0875-G95Y 6.25Y 8.5/13 0.64
12 E 51 Green-yellow 0963-G66Y 2.50Y8/10 0.60
12 E 53 Green-yellow l266-G45Y 5GY 7/11 0.44
14 E 51 Green 1854-GO9Y 2.5G6.5/8 0.34
14 F. 53 Green 2854-G5G 5G 5/10 0.22
16 E 53 Blue-green 3049-B5OG 7.5BG 5/8 0.22
18 E 49 Blue 0710-B64G SB 9/2 0.79
18 F. 50 Blue 0822-BuG 7.5B 8/4 0.60
18 E 51 Blue 1847-BO6G 7.5B 6/8 0.31
18 E 53 Blue 2959-B lOB 4/10 0.15
20 F. 5 I Purple-blue I548-R89B SPB 6/10 0.32
00 E 53 Black 9500-N N 1.5 0.05
00 F. 55 White 0000-N N 9.5 0.85
20 Societyof Light and Lighting and National Physical Laboratory

4.5 RAL Design System

A further system commonly used by designers is the RAL Design System which was
developed by the German Institute for Quality Assurance and Specification about 70
years ago. it is usedto define surface colours forpaint and othermanufactured materials
and is based on an atlas of 1688 colours. The method ofdefining colours is similarto the
Munsell systemusing three parameters for each colour, i.e. hue (red, yellow etc.), light-
ness (which is similar to the Munsell value) and chroma (similar to Munsell chroma).
Whilst the earliersystems described were based on visual organisation, the RAL system
is based on a mathematical colour space using the CIE chrornaticity system. The RAL
coding system uses threenumbergroupswhich give the positionin the CIEcolourspace
(CIELab using either cartesian or cylindrical co-ordinates) and in the RAL atlas.
(Web site: http://www.ral.dc.

5 Measurement of reflectance and colour

There are several methods by which to measure the reflectance and/or colour of a sur-
face, and these are described below. Whichever method is chosen, it is important to
consider the magnitude and importance of the measurement errors which can arise. To
minimise the magnitude of these errors, instruments should be calibrated or checked
using reflectance or colour standards which have themselves been calibrated to an ap-
propriatedegree ofaccuracy. For example, a manufacturer oftextiles or paints will usu-
ally use a spectrophotometer for making measurements and check its performance using
reference materials calibrated at NPL (the National Standards Laboratory in the UK) or
other approved laboratory. A householder repainting the woodwork in a room, on the
other hand, will generally make visual comparisons between the existing paint and a
swatch ofcolour samples from a paint manufacturer; accuracy is not an important con-
sideration in this process. Further information on reflectance and colour standards is
given in Section 6.

5.1 Measurement using a spectrophotometer

The mostaccuratemeasurements ofthe reflectance and colour of a surfaceare made in
the laboratory using a specially designed measurement system called a spectrophotom-
eter. A spectrophotometer typically comprises a light source, a monochromator or other
means of splitting the light into its constituent wavelengths, a compartment or holderin
which the reflectance sample is placed, and a detector. For a specular surface it is only
necessary to arrange the detectorat the correct angle to the sample so as to collect the
reflected beam. For a diffusing material, however, some form of optical arrangement
(which could include lenses, mirrors, a small integrating sphere etc.) is needed between
the sample and the detectorso as to collectall the lightreflected regardless ofangle,orto
collect the light reflected into a defined range of angles. It is important to note that for
some materials the reflection properties also vary with the angleofthe incidentlight. To
avoid the confusion which may arise from the multiplicity ofdifferent geometries under
whichmeasurements couldbe made, a numberofstandard measurement conditions have
been defined.
Those often used include:
0°/i' = light incidentperpendicular to the surface, all reflected light collected
Lighting Guide Surface reflectanceand colour 21

O°/d = lightincidentperpendicular to the surface, diffused lightcollected, specular

reflection excluded
O°/45° = light incident perpendicular to the surface, light reflected at 45° to the
normalcollected
8°/i'= light incidentat80 to the perpendicular to the surface, all reflected light
collected
These standard geometrical conditions are useful when comparing the reflectance of
surfaces with similar light-scattering properties. For a more complete description ofthe
reflectance of a surface (as required, for example, in the design of road lighting) meas-
urements may be made at a range of incidentand reflectedangles.
Their complex nature means that spectrophotometers are relatively expensive and
difficult-to-use instruments and their use is therefore generally restrictedto those appli-
cations where accurate and precise measurements of reflectance and colour are needed,
such as the manufacture ofpaints and textiles.

5.2 Measurement using a colorimeter

For many applications it is the colour of a surface which is of interest, rather than its
detailed reflectance properties. As mentioned above, colour depends not only on the
properties ofthe surface but also the characteristics ofthe light falling on it. A measure-
ment using a spectrophotometer allows the colour ofan object to be calculated for any
known illuminating source. Other instruments, called colorimeters, can measure colour
directly using detectors which are designed to mimic the response of the tristimulus
distributions for the ClE 193 1 Standard Observer. For the results from such instruments
to be meaningful, it is important that the illuminating source used during the measure-
ments is the same as that to be used in the final environment. For specification and
measurement purposes, one ofa range of agreed standard illuminants is generally used,
the most common being simulated 'D65' (an approximation to daylight with a correlated
colourtemperature of 6500 K) and 'A' (a tungsten filament lampat a colour temperature
of2856 K) (CIE, 1986).
As in the case ofspectrophotometric measurements, the results ofmeasurements made
using a colorimeter will depend on the geometry of illumination and collection of the
light. For simplicity, most colorimeters for measuringthe colour ofa surfaceuse a fixed
geometry, typically 0°/45°, and therefore the measuredcolour may not truly represent
the visual appearance of a surface, particularly if it has strong specular reflection or is
pearlescent (i.e. its colour varies with angle ofview).
Some colorimeters can also be used to measure the average reflectance ofa surface.
Once again, the results depend on the illurninant and measurement geometry used.
The level of uncertainty associated with measurements made using a colorimeter is
generally considerably higher (i.e. worse) than that for spectrophotometric measure-
ments, due to the fact that the detectors never perfectly match the tristimulus distribu-
tions. This means that their value for making absolute measurements (as required, for
example, in the design ofa material to match a numerical colourspecification) is rather
limited. However they are ideally suited to the evaluation of small colour differences
between surfaces and they find widespread use in applications such as the quality con-
trol of paints and dyes, where a production sample is compared with a reference sample
for quality control purposes.
22 Societyof Light and Lighting and National Physical Laboratory

5.3 Measurement using luminance and llluminance meters

For a perfectly diffusing surface, the luminance L (in cd/rn2)is relatedto the reflectance,
rand the illuminance E(in lux) as follows:
L= r. E / it
Therefore if one measures the luminance of such a surface and the illuminance of the
light falling onto it, the reflectance can be determined.
For the purposes oflightingdesigncalculations, it is frequentlyacceptable to assume
that all the surfaces within a space act as perfectdiffusers. This relationship then gives a
method by which to calculate the reflectance of surfaces in the field by the use of a
luminance meter and an illuminance meter. Howeverit should be notedthat the relation-
ship does not hold for heavily textured surfaces, or those with significant specular re-
flection. For these surfaces the relationship between luminance and illuminance is a
complex one; this method will give erroneous results and thereforeshould not be used.
It is important to use meters that have been calibratedfor the type of lamp which is
beingusedto illuminate the surface, preferablyby an accreditedlaboratory (BS667:1996
andBS 7920: 1998). The procedure is as follows:
(i) Use the luminance meter to measure the luminance ofa representative portion ofthe
surface, taking care to align the meter so that its measurement axis is perpendicular to
the surface and ensuring that it does not significantly affect the illumination falling on
the surface. (The majority of surfaces approximate a perfect diffuser most closely at
angles close to the perpendicular.)
(ii) Place the illuminance meter on the surface at the positionpreviously viewed using
the luminance meter and record the illuminance.
(iii) From these two measurements, calculate the reflectance usingthe relationship given
above.
There will typically be a fairly large uncertainty associatedwith the result of such a
measurement, even in the case ofa perfect diffuser, andthus for critical applicationsmeas-
urement in a laboratory using one of the techniques described above may be necessary.

5.4 Field measurements using the reflectance card

The reflectance sample card provided with thispublication is intended to enablelighting
designers and architects to make an assessment of surface reflectance to the degree of
accuracy appropriate for use in lighting design software or manualcalculation routines.
The following procedure is recommended:
(i) Hold the card against the surface, at a comfortable viewingdistance, taking care not
to obstructthe light falling on to the surface and ensuring that the card is uniformly
illuminated with diffuse light.
(ii) Select the column whichmost closely resembles the colourofthe surface. In situa-
tions where there is doubt over which column is closest in appearance, use the two (or
more) which are most similar and take the average result. In additionthe grey column
can be used with any coloursurface.
(iii) Select the sample(s) within the chosen column(s) which appear to be closest in
brightness to the surface (which can be viewed throughthe holes in the card). Read off
the reflectance value(s)for the type oflampwith whichthe surface is beingilluminated.
Where the brightnessof the surface appears to lie between two adjacent samplesin a
colour column, the mean reflectance of the two samplesmay be used.
Lighting Guide. Surface reflectanceand colour 23

(iv) If necessary take the mean ofthe reflectance values from the two (or more) colour
columns chosen.

5.5 Visual assessment

The human eye is a very adaptable organ, dealing easily with changes in illumination
rangingfrom moonlightto brightsunlight. However this adaptability makes it relatively
poor at making numerical assessments of illumination levels, colour, reflectance etc.
Visual 'measurements'can only be made by making comparisons between a surface
which has been measuredusing a physical measurement device and an unknown sur-
face. A typical example is colour matching ofpaints or textiles,where visual compari-
Sons betweenthe test and reference material are an excellentmethod by which to judge
when an acceptable match is obtained. Indeed even the best colorimeters and
spectrophotometers find it difficult to exceed the eye's performance in the identification
of small differences in reflectance or colour.
As in the case of measurements using colorimeters, luminance and illuminance me-
ters, or the reflectance sample card, the results of a visual assessment ofa surface will
depend on the illuminating source and the measurement geometry. It is not unusual for
two surfaces to look similaror to co-ordinate with each other under one illuminant, but
not under another— an effect knownas metamerism. Similarly a glossy and a matt sur-
face are compared, the reflectance ofthe two may appearsimilar at some angles but very
differentat others.

6 Reflectance and colour standards

The reflectanceand colour of a surface, whetherdiffuse or specular, can be measured
using a number oftechniques, as described in Section 5. It is very important to many
industries that measurements made on a wide range of spectrophotometers and
colorimeters agree. Colourmatching and co-ordination is very important in many appli-
cations and disagreements between instrument scaleswould make this almost impossi-
ble.
In order to achieve agreement, reflectance and colourtransferstandards are available,
traceable to NPL (the UK's National Standards Laboratory). There are wide range of
standards available includinga set of 12 ceramic colour standards, black,whiteand grey
tiles,and white, grey and coloured Spectralon (a form ofscinteredPTFE). Some ofthese
are available with both glossy and matt surfaces. The white, grey and black reflectance
standards are used to check the linearity of the detectorsystem and scale i.e. to check
whetherthe instrumental readings ofreflectance are correct. Holmiumoxide reflectance
tiles are used to check the wavelength scale i.e. to confirm whetherthe wavelength dis-
played by an instrument is correct. Ceramic colourstandards are used to checkthe over-
all performance of instruments and their software for the measurement of reflectance
and colour. Figures 12 and 13 show typical reflectance as a function of wavelength for
grey and coloured ceramic tiles. Figure 14 shows a selection of the standards available
from NPL.
Some laboratories use pressed barium sulphate or magnesium oxideas a whitediffuse
reflectance standard, but these can be very easily damaged and have a short lifetime
before degradation, so are not recommended. Other materials such as opal glass are also
sometimes used as standards, but they are more susceptible to damage than ceramic tiles
or Spectralon standards and their use is thereforedecreasing.
24 Societyof Light and Lighting and National Physical Laboratory

80

Pale grey
60

U
C

Mid grey

Deep grey

0- 400 500 600 700

Wavelength, nm

Figure 12 Diffuse reflectanceas a function of wavelength for pale, mid and deep grey tiles.

80

60

40
20

0
400 500 600 700
Wavelength, nm

Figure 13, Diffuse reflectanceas a function of wavelength for ceramic colour standards

7 Surface appearance
Much of this document has been concerned with the physical parameters of surface
materials andtheirrelationship to lighting andparticularly reflected light. However these
do not definewith anycertainty the exact appearance ofthe surface. It is notthe intention
to discussthiscomplex subject in any real detail butto ignore it would be equally inappro-
priate. A surface will appear as it doesbecause ofthe physical parameters olthe surface,
Lighting Guide: Surface reflectanceand colour 25

Figure 14. Selectionof co/our standards

the way it is illuminated and its relationship to the rest ofthe scene as well as the visual
state of the observer.
The physical attributes ofthe surface that need to be considered are its reflectance, its
surface colour and the surface finish includingthe structure or make-up ofthe material
which could affect its appearance. Regarding the illumination the level of illuminance,
the illuminance distribution overthe surface andthe spectral distribution ofthe lightwill
all need to be considered. The luminance and colour composition of the surrounding
scene will also have some effect, whichin turn will affect the observer'svisual state and
particularly the brightnessand colouradaptation. Brightness and colour constancy may
also have an effect, but only ifit breaksdown, whichexcept in rare cases this is unlikely
to happen in real spaces. Constancy will normallybe upheld as long as the observer
receives sufficientvisual cues as to the way the scene is lit and for the purposes of this
document it will be presumed that this is the case.

Considering each ofthe elements in turn:

Surface reflectance: this is the feature of a surface that determines its degree of light-
ness or darkness. For example ifit is requiredthat a surface,or an interior, appears 'light'
then it will be essential for the surface or surfaces to have a high reflectance. A dark
surface can never be made to appear 'light' how ever much light is incidentupon it.
Conversely a high reflectance surface will never appear truly dark just by using a low
level ofilluminance - it will only appear dingy or gloomy.

Surface colour: this is the feature of a surface that determines its unique colour i.e. its
redness, blueness, greenness or yellowness and as long as there is a sufficient illumi-
nance and an amount of neutral or achromatic surface colours within the field of view
then the appearance should be largelycorrect.

Surface finish and texture: this will not affect the total amount of lightbeingreflected
from the surface but it will affect its appearance. For example a truly matt surface will
have the same brightness from any position of view although its projected area may
change. However a surface with some degree of polish will have two elements in its
appearance. It will have that lightwhichpenetrates the polishedskin ofthe surface and
26 Societyof Light and Lighting and National Physical Laboratory

Figure 15. Polishedsurface illuminated to show colour and highlight.

Figure 16. Texturedsurface illuminated by diffuse, even illumination.

Figure 17. Textured surface lit with directional, oblique light

Lighting Guide. Surface reflectanceand colour 27

reflects from the base colourandemerges fromthe surface as showing the surface as that
colour. On top of that will be a reflected image ofthe illumination source providing the
'gloss' to the surface. The size andposition ofthis will dependon the geometric relation-
ship between the illumination source and the observer. For textured surfaces, ifthe sur-
face is lit very uniformly from all directions then the surface will appear as a uniform
reflectance see Figure 15. Butifthe illuminant has a prominent direction relative to the
surface then the surface will take on a very different appearance see Figure 16. This
will be because the lightwill reflect from the peaks ofthe surface, as a luminance deter-
mined by the base reflectance ofthe material, but therewill also be shadows createdby
those peaks which will cause the surface to have a generally lower overall brightness. It
is the patternofluminances withinthe surface that providesthe information to the brain
that the surfaceis texturedand notjust a surface with a lower brightness—see Figure 17.
A textured surface, particularly when lit obliquely, will create a greater visual interest
withinthe scene than the overall brightness would suggest.

Surface illumination: this divides into threemain considerations.

First is the spectral powerdistribution ofthe light source, whichwill affect the colour
appearance ofthe surface. For example if a surface is green in colour and it is lit with a
source that contains radiation of all wavelengths then the surface will appear green. If
however it is lit by a source that has no greenradiation e.g. a low pressure sodium lamp,
then the surface will appear grey or black. If it is required that surface colours should
appear reasonably accurate then the illuminant must have a CIEColourRendering Index
(R) ofnot less than 80 and ifhigh fidelity colourappearance is required then theRa must
be not less than 90. Lighting a surface with a coloured illuminantis usually only done
when a special effect is required for decorative purposes e.g. exterior floodlighting. In
this case great accuracy is rarely a requirement.
Secondly, the level olsurface illuminance will also need to be considered. For exam-
ple if a painted mural on a wall was lit at a low illuminance, e.g. lower than 50 lux, the
colours would start to appear indistinct and accurate colour discrimination could begin
to be impaired. Perhaps more importantly the colours would begin to loose their vi-
brancy. Ifthe full colour impact ofa surface is required then an illuminance of around
200 lux or above will be required. The ultimate effect will also be determined, to some
extent, by the observer's level of adaptation. For example if a room is lit at a very low
general level of illuminance and hence the occupant's adaptation level is low, then a
surfacelit to say 50 lux couldbe acceptable. This is the situation commonly found in art
galleries, used for the display of light sensitive exhibits e.g. watercolours.
Thirdly the illuminancc distribution should be considered. For example ifa surface is
lit with a high illuminance uniformity then it will appear uniformly brightand uniformly
coloured. Alternatively if the surface is lit very non-unifonnly, e.g. with an illuminance
diversity (minimum / maximum) ofsay I : 100 or greater, then the surfacewill take on an
unevenappearance both in termsofbrightnessandcolour; there are situations where this
effectmay be acceptable or even required. The question the designer needs to ask is what
is the effect that is required? The surface finish will also have some effect on whethera
surface appears evenly lit or not. For example ifa surfaceis heavily textured then a high
illuminance diversity will often be masked, becomingacceptable or even preferable in
terms ofits lit appearance.

ElTect of contrastbetween one surface and another:the apparentbrightness and col-

our appearance of a surface will be influenced by the degree of contrast between the
28

S Societyof Light and Lighting and National Physical Laboratory

Figure 18. Assimilation. When viewed fromabout 2 m the uniform grey background looks
lighterwith whitejointing, darker with black jointing.Although assimilation is predominantly an
aspect of lightening or darkening, it is not exclusivelyso. Thussmall scale areas surrounded by
narrow, distinbctively coloured joints, say red or blue, may appear tingedwith the same hue.

surface and adjacent or surrounding surfaces. For example ifan achromatic surface of a
particular luminance is surrounded by a surface of a lower luminance the tendencywill
be for the surface to appear brighter and conversely ifthe surround is at a higher lumi-
nance then the surface will appear darker. The effect is known as brightness induction
and has the effect of exaggerating a luminance difference between one surface and an-
other se Figure 18.
A similareffectoccurs between coloured surfaces and neutral surfaces, and between
different coloured surfaces. A coloured surface will take on a hint ofthe complementary
colour ofthe surrounding surface see Figure 19. This effect is known as colour induc-
tion. Therefore a neutral grey surface ifsurrounded by a green surface will takeon a red
or pinkish tingeand the same grey surface ifsurrounded by a blue surface will take on a
yellowish tinge and vice versa. This means that a greater contrast can be achieved by
using this effect.

Visual adaptation (brightness and colour): an element of the human visual system
includes an automatic processknown as adaptation. In the case ofbrightness adaptation
it works rather like the automatic exposure control on a camera in that it adjusts the
exposure value to the average luminance ofthe scene and the areas ofhigher luminance
appearlight whilst the areasoflower luminance appear dark. This means that a particu-
lar luminance will vary in brightness if the adaptation state changes. A similar process
occurs for colour although this is usually less noticeable. If a person works for some
time in a room lit with a particular coloured light source, for example in a photographic
darkroom lit with red light. After a time the person will cease to be aware of the red
source dueto colouradaptation, but when switching to a normal 'white' source the scene
will take on a greenish tinge for a short time until the eye readapts. The process of
adaptation takestime to change from one level to another and designers need to be aware
ofthis andto engineerthe lighting to allowthis to happenwithoutloss ofvisual comfort
or visibility. This will be required, for example, in a museum where adjacent galleries
are lit to different levels for conservation purposes. To allowthe transition to takeplace
a linking corridorshould be provided which is lit at an intermediary luminance.
Lighting Guide: Surface reflectanceand colour 29

Figure 19. Contrast and assimilation effects. The small squareswithin the patternare printed the
same grey as the separatesquare but look different, through the combined effects of contrast
and assimilation.

Theabovephenomena will all affect the appearance of surfaces, some more than others,
but they all need to be borne in mind by designers when planning the appearance of a
task and the appearance of a lit environment to achieve the most effective performance
and visual quality.

8 Explanation of terms
This Guide uses a number of terms, both scientific and descriptive, that may not be
familiarto all readers. The following is provided to help overcome the problem.

Area-weighted reflectance: This represents the average reflectance of a surface which

is made up of areas ofdifferent reflectance but weighted by the areas ofthe individual
parts.

Colour rendering: A general expression for the appearance of surface colours when
illuminated by light from a given illuminant compared, consciously or unconsciously,
with their appearance under a reference illuminant. Goodcolourrendering impliessimi-
larity of appearance to that under an acceptable light source, such as daylight.

Colour rendering index: A measure of the degree to which the colour appearance of
surfaces illuminated by a given lightsource conform to those ofthe same surfaces under
a reference illuminant, suitable allowance having been made for the state of chromatic
adaptation. The system commonly used is that adoptedby the CIE in 1965 (updated in
1974 and 1994) which provides a measure of difference betweenthe two values.

Colour temperature(kelvin):The temperature of a Planckian, or 'full radiator' whose

radiation has the same chromaticity as that ofthe light source being considered.

Correlated colour temperature(kelvin): The temperature of a 'full radiator' which

emits radiation with a chromaticity nearest to that ofthe light source being considered.
30 Societyo Light and Lighting and National Physical Laboratory

Diffuse reflection: Reflection in which the reflected light is diffused and there is no
significant specular reflection, as from a matt paint. See also Uniform or Lambertian
Diffuser.

Gloss: The mode of appearance by which the reflected highlights of objects are per-
ceived as superimposed on the surface due to the directionally selective properties of
that surface.

Gloss factor: Ameasure ofthe specular reflectance ofa surface, usuallymeasured at an

angle of 45°. Measurementsare generally made using a glossmeter (a spectrophotom-
eter or colorimeter with a fixed angle of illumination and view) by reference to a cali-
brated gloss standard (typically a polished clear glass plate with blackenedbase and
sides, which has a gloss factor of—lOO).

Illuminance (lumen/square metre): The luminous flux density at a surface, i.e. the
luminous flux incident per unit area.

Luminous flux (lumen): The light emitted by a source, or receivedby a surface. The
quantity is derived from radiant flux by modifying the radiation in accordance with the
spectral sensitivity of the standard eye as described by the CIE standard photopic ob-
server (V(?)).

Luminance (candela/square metre): The physical measure ofthe stimulus whichpro-

duces the sensation ofbrightness measured by the luminous intensity ofthe light emitted
or reflected in a givendirection from a surface element, divided by the projected areaof
the element in the same direction. The relationship between luminance and illuminance
for a reflecting surface is given by the equation:

luminance = illuminance >< reflectance factor

Note: for diffuse surfaces reflectance factoris replacedby reflectance.

Luminance factor: The ratio ofthe luminance ofa reflecting surface, viewedin a given
direction, to that ofan identically illuminated 'perfect white' uniformdiffusing surface.
For a non-matt surface the luminance factormay be greateror less than the reflectance.

Mixed reflection: Partly specular and partly diffuse reflection, as from smooth, glossy
paint. See Luminance factor.

Radiant flux (watt): The power emitted, transferred or receivedas radiation.

Reflectance:The ratio ofthe luminous fluxreflected from a surface to the luminous flux
incident on it. Except for rnatt surfaces, reflectance will depend on how the surface is
illuminated but especially on the direction of the incidentlightand its spectral distribu-
tion. The valueis always less than unityandis expressed either as a decimal or a percent-
age.
Note: Reflectance can also be measuredbased on radiantflux ratherthan luminous flux.
Lighting Guide. Surface reflectanceand colour 31

Reflectance factor: The ratioofthe radiant or luminous flux reflected into a given cone
to that reflected in the same cone by a perfectreflecting diffuser, identically irradiated or
illuminated.

Spectral (power or energy) distribution: Sometimes referredto as 'spectralcomposi-

tion'. The variation of radiant power (or energy) over a range ofwavelengths.

Specularreflection: Reflection withoutdiffusion in accordance with the laws ofoptical

reflection, as in a mirror.

Uniform (or Lambertian) diffuser: An ideal surface for which the radiation coming
from that surface is distributed angularly according to Lambert's cosine law (i.e. whose
radianceor luminance is the same in all directions andwhose luminous intensity at angle
0 to the normal to the surface varies as the cosineofthe angle0) irrespective to how it is
illuminated.

Further reading
CIBSE Codefor interiorlighting (London: CIBSE) (1994)
CIE Publication 15.2 (1986): Colorimetry (Vienna, Central Bureau ofthe CIE) (1986)
BS 667: 1996: Specification for illuminance meters (London: British Standards Institu-
tion) (1996)
BS 4800: 1989 (1994): Specification for paint colours for building purposes (London:
British Standards Institution) (1989)
BS 4900: 1976 (1994): Specification for vitreous enamel colours for building purposes
(London: British Standards Institution) (1976)
BS 4901: 1976 (1986): Specification for plastics colours for building purposes (London:
British Standards Institution) (1976)
BS 4904: 1978 (1997): Specification for external cladding colours for building purposes
(London: British Standards Institution) (1978)
BS 5252: 1976: Framework for colour co-ordination for building purposes (London:
British Standards Institution) (1976)
BS 6770: 1988 (1994): Guide for the exterior colours for park homes (mobile homes),
holiday caravans and transportable accommodation units (London: British Standards
Institution) (1988)
BS 7920: 1998: Specification for luminance meters (London: British Standards Institu-
tion) (1998)
Bean, A R and Simons R H, Lighting engineering: applied calculations (Oxford:
Butterworth-Heinemann) (2001)
Coaton J R and Marsden A M, Lamps and lighting 4th edn, (London: Arnold) (1997)
Hard, A. and Sivik, L, ColorRes. AppI., 6, 129 (1981)
Hunt, R.W.G. Measuring colour 2nd edn (Chichester: Ellis Horwood) (1991)
Hunter R S & Harold ft W, Themeasurement ofappearance (Chichester: Wiley& Sons)
(1987)
Nickerson, D, Coloi Res. Appi. 1, 7, 69 and 121 (1976)
Padgham C A& Saunders J E, ThePerceptionoflightandcolour (London: Bell& Sons)
(1975)
32 Societyof Light and Lighting and National Physical Laboratory

Reflectance sample card

The reflectance sample card is intended to enable lighting designers and architects to
makevisualassessments ofsurfacereflectance to the degree ofaccuracy appropriate for
use in lighting design software or manual calculation routines.

Table of surface reflectancevalues

Light source A B C D E F C H
grey brown ochre olive red blue green yellow
Row I Incandescent 10 14 25 18 25 15 13 39
Fluorescent 840 10 13 25 18 23 16 15 38
Daylight D65 10 13 23 17 19 17 16 36
Row 2 Incandescent 15 18 31 25 31 24 21 47
Fluorescent 840 15 18 31 25 29 25 23 47
Daylight D65 IS 17 30 24 25 26 24 45
Row3 Incandescent 24 27 40 33 40 35 34 56
Fluorescent 840 24 27 39 33 38 35 35 56
DaylightD65 24 26 38 33 35 37 36 54
Row 4 Incandescent 36 41 50 47 53 49 47 64
Fluorescent 840 36 41 50 47 52 50 48 64
Daylight D65 36 40 48 46 49 51 50 63
Row 5 Incandescent 55 58 64 61 65 63 62 73
Fluorescent 840 54 58 64 61 64 63 62 73
DaylightD65 55 58 63 61 63 64 64 72
Row6 Incandescent 74 77 79 77 77 78 78 81
Fluorescent840 73 76 79 77 76 78 78 81
DaylightD65 74 77 79 77 76 79 78 81

Procedure
I Holdthe cardagainst thesurface, at a comfortable viewing distance, takingcarenot to obstruct
the light falling onthe surfaceand ensuring thatthe cardis uniformlyilluminated with diffuse
light.
2 Selectthe column whichmost closely resembles the colourofthe surface. In situations where
there is doubt over which column is closest in appearance, use the two (or more) which are
most similar and take the average result. In addition, the grey column can be used with a
surfaceof any colour.
3 Selectthe sample(s) within the chosencolumn(s) which appearto be closestin brightnessto
the surface(which can be viewedthrough the holes in the card). Read off the reflectance
value(s) for the type of light source with which the surface is being illuminated. Wherethe
brightness ofthe surface appearsto lie betweentwo adjacentsamplesin a colourcolumn, the
mean reflectance of the two samples may be used.
4 Ifnecessary, take the mean of the reflectance values from the two (or more) colour columns
chosen.
Note: Although every care has been taken to select stable printing inks for the reflectance
sample card, the patches are likelyto change in colour and/or reflectance ifthey are exposed to
light for longperiods. Users areadvisedto store the cardin theprotectivesleeve provided when
it is notin use. This will also help to keep the cardclean.
In the event that the reflectance sample card becomes dirtyor damaged, a replacement copy
can be purchased from CIBSE Publications.