Cultura e Scienza del Colore - Color Culture and Science


 4503/15 | Cultura e Scienza del Colore - Color Culture and Science

1Laura Bellia 
bellia@unina.it
1Nicola Agresta 
1Luigi Rossi
1Gennaro Spada

1 DETEC, Università degli Studi di 
Napoli “Federico II”

Measured color changes 
under fluorescent and LED lamps 

1. INTRODUCTION

In the field of electric lighting, the increasing 
popularity and spread of the LED sources, 
together with the availability of traditional lamps 
with different spectral characteristics, compels 
the designer to consider in his choices not 
only aspects related to energy saving, but also 
to take into account lighting quality, meaning 
for it a correct and pleasant perception of the 
surrounding environment [1],[2],[3].

Once fixed the values of illuminance and 
luminance distributions, the parameters that 
characterize the sources regarding quality 
of light, according to European Standard EN 
12464-1[4], and that strongly influence the 
perception of objects and colors are the color 
appearance, identified by Correlated Color 
Temperature (CCT), and Color Rendering Index 
(CRI). However in presence of sources with 
similar CCT and CRI but with different Spectral 
Power Distributions (SPD), different perceptual 
effects occur [5],[6],[7]. 

In this paper, the effects of two kind of sources, 
fluorescents and LEDs, both with high CRI, 
are examined by means of color measures on 
different objects.

2. EXPERIMENTAL SET-UP
2.1. DESCRIPTION OF THE TEST ROOM AND            
       LIGHT SOURCES

The test room is quite 3.00m x 3.00m x 3.00m 
with completely blank walls (ρ

M
 = 70%) and a 

false ceiling with controlled lights. The furniture 
consists in a desk with a chair centrally located 
in respect to the lights. Luminaires consist in 
2 fluorescent and 4 LED sources. Technical 
characteristics of each LED and fluorescent 
source provided by manufacturer are shown in 
the Table 1.

Fluorescent LED

Lamp 4xT16/24W 1x1824/27W

Luminous 
flux

7000 lm 1824 lm

Efficacy 49 lm/W 64 lm/W

CRI 1B (80<Ra<90) 90-91

For the characterization of the sources, a 
Konica-Minolta CS2000 spectroradiometer was 
used; the instrument was pointed towards the 
lamps to obtain the SPDs.

2.2. DESCRIPTION OF LIGHT SCENES

Lights are connected to Luxmate Emotion 
information system by Zumtobel. Thanks to 
this device, it was possible to dim both lighting 
emissions and change color temperature, in 
order to obtain 8 light scenes with similar 
illuminance values on the desk, measured 
with 2 Konica-Minolta T10 lux meters centrally 

Figure 1 – Test room
Figure 2 – View of the false ceiling                                 

Table 2 – Light scenes set-up

Table 1 – Technical characteristics of 
luminaires                                

CCT [K] Dim [%] E
m 

[lux]

Fluo 
2700K 2924 100 653

Fluo 
3000K 3011 93 549

Fluo 
4000K 4178 94 558

Fluo 
6500K 7100 100 621

LED 
2900K 2940 97 634

LED 
3150K 3020 94 545

LED 
4200K

4199 95 548

LED 
6100K

7195 100 628



46 Cultura e Scienza del Colore - Color Culture and Science | 03/15  

located in respect to the lights.

The following analyses have shown that light 
scenes with intermediate CCT (Fluo 3000K, Fluo 
4000K, LED 3150K and LED 4200K) reported 
results less interesting than the other scenes: so 
only the results of Fluo 2700K and 6500K, LED 
2900K and 6100K will be shown. As in Table 2, 
Fluo 2700K and LED 2900K scenes have nearly 
the same measured CCT, as also Fluo 6500K 
and LED 6100K: since now, “warm lights” will be 
referred to first ones while “cold lights” will be 
referred to second ones.

2.3. DESCRIPTION OF COLORED OBJECTS

On the desk 6 different colored objects (LEGO® 

blocks) were placed: blue, light blue, green, 
yellow, orange, red. For each one of them, 
spectral reflectance and color coordinates under  
illuminant D65 were measured.

3. ANALYSIS AND RESULTS
3.1. SPECTRAL DISTRIBUTION                         
       OF REFLECTED LIGHT ANALYSIS 

Spectral reflectance of each colored object was 
measured with the spectrophotometer; lamps’ 
SPDs were obtained with the spectroradiometer 
while spectral distribution of reflected light 
was obtained pointing the spectroradiometer 
on each colored object under each light scene. 
Results are reported in figures 5 – 10 in which 
not all but just the most interesting cases were 
represented. 

For each figure, diagrams corresponding to 
(C) show clearly that they are the result of 
the combination of the previous two: where 
the block’s spectral reflectance increases 
or  decreases the value of emitted light for 
those wavelengths in the reflection diagram is 

Table 3 – Color coordinates of  
LEGO® blocks

      

      

Figure 3 – SPD Fluorescent 2700K 
(A), 3000K (B), 4000K (C) and 
6500K (D).

Figure 4 –  SPD LED 2900K (A), 
3150K (B), 4200K (C) and 6100K 
(D).

Figure 5 – Blue block spectral 
reflectance (A), LED 2900K SPD (B) 
and spectral distribution of reflected 
light (C). According to the spectral 
reflectance, in all light scenes red 
component is reduced while blue one 
is maximized, even with warm lights as 
shown in the example.
Figure 6 – Light blue block spectral 
reflectance (A), Fluo 6500K SPD (B) 
and spectral distribution of reflected 
light (C). Spectral reflectance shows 
high values of both blue and red 
component, increasing block’s 
reflection values in low and high 
wavelengths.
Figure 7 – Green block spectral 
reflectance (A), LED 2900K SPD (B) 
and spectral distribution of reflected 
light (C). Green component of all 
lights’ SPDs is maximized while blue 
and red ones are reduced, especially 
with LEDs.

Specular component

included excluded

L* a* b* L* a* b*

Blue 45.68 -7.48 -41.96 42.35 -8.55 -44.87

Light blue 63.76 -11.01 -27.64 61.34 -11.79 -28.95

Green 60.43 -51.72 36.90 57.98 -56.32 40.75

Yellow 80.84 8.48 78.32 79.15 8.65 85.39

Orange 66.60 45.37 55.38 63.63 47.94 65.07

Red 42.93 50.12 32.00 30.09 55.06 40.00

    

    

    



 4703/15 | Cultura e Scienza del Colore - Color Culture and Science

Figure 8 – Yellow block spectral 
reflectance (A), LED 6100K SPD (B) 
and spectral distribution of reflected 
light (C). In these cases, the only 
notable change between lamps’ 
SPDs and reflection diagrams is the 
reduction of the blue component.

Figure 9 – Orange block spectral 
reflectance (A), LED 6100K SPD (B) 
and spectral distribution of reflected 
light (C). As for the yellow block, in 
the reflection diagrams it’s just the 
blue component of lamps’ SPDs to 
be reduced.

Figure 10 – Red block spectral 
reflectance (A), Fluo 2700K SPD (B) 
and spectral distribution of reflected 
light (C). The red block case is similar 
to the yellow and orange ones.

    

    

    

amplified or reduced.

 As general rule, we can say that: blue prefers 
cold lights; light blue is better under cold lights 
but it doesn’t change too much with the other 
light scenes; green prefers cold lights; yellow 
is better under warm lights; orange and red are 
good under every kind of light.

3.2. CCT AND TECHNOLOGIES                       
       COMPARISONS  

After that, for each color two different 
comparisons were made: Correlated Color 
Temperature with the same technology and 
technology comparison with the same CCT. This 
analysis helped to find the best and the worst 
kind of light for each color.

Figure 11 – Fluo 6500K lamp has a 
wide range of blue wavelenghts while 
Fluo 2700K lamp have peaks in other 
colours; LEDs have quite the same 
trend but with higher values for blue 
and lower for red compared to the 
fluorescent sources. Cold lights are 
preferred, with LED 6100K as best 
source and Fluo 2700K as worst.

CCT comparison for blue

Technologies comparison for blue

      Fluo 2700K            Fluo 6500K            LED 2900K           LED 6100K  

Figure12 – As for blue, in light blue 
reflection diagrams fluorescent sourc-
es have other colour components 
while LEDs don’t have the previous 
trend, with LED 2900K having a 
peak in red. Technologies comparison 
shows how all the lights cover a wide 
range of wavelenghts with a prefer-
ence for LED 6100K.

CCT comparison for light blue

Technologies comparison for light blue

      Fluo 2700K            Fluo 6500K            LED 2900K           LED 6100K  



48 Cultura e Scienza del Colore - Color Culture and Science | 03/15  

Figure 13 – CCT comparison shows 
how lights with the same technology 
have quite the same trend for green. 
However it’s possible to observe 
how cold lights have more green 
component than warm lights that 
increase red component. In this case 
Fluo 6500K has a peak in green 
wavelenghts more than LED as in 
previous cases

Figure 14 – Considerations made 
about green reflection diagrams 
could be applied to yellow’s too: 
CCT comparison shows how lights 
with the same technology have quite 
the same trend, with a preference for 
warm lights. In particular, technologies 
comparison shows how similar is the 
trend of the two sources.

Figure 15 – CCT comparison shows 
how lights with the same technology 
have quite the same trend for orange 
reflection but it’s possible to see 
how cold lights have a wider range 
of wavelenghts than warm lights, in 
which orange and red component is 
maximized.

Figure 16 – Considerations made 
about orange reflection diagrams 
could be applied to red’s: in this case 
too, warm lights should be preferred 
to cold ones.

CCT comparison for green

Technologies comparison for green

      Fluo 2700K            Fluo 6500K            LED 2900K           LED 6100K  

CCT comparison for yellow

Technologies comparison for yellow

      Fluo 2700K            Fluo 6500K            LED 2900K           LED 6100K  

CCT comparison for orange

Technologies comparison for orange

      Fluo 2700K            Fluo 6500K            LED 2900K           LED 6100K  

CCT comparison for red

Technologies comparison for red

      Fluo 2700K            Fluo 6500K            LED 2900K           LED 6100K  



 4903/15 | Cultura e Scienza del Colore - Color Culture and Science

CCT comparison for orange

Technologies comparison for orange

      Fluo 2700K            Fluo 6500K            LED 2900K           LED 6100K  

CCT comparison for red

Technologies comparison for red

      Fluo 2700K            Fluo 6500K            LED 2900K           LED 6100K  

3.3. CHANGES OF CHROMATICITY               
       COORDINATES                                       
       UNDER DIFFERENT LIGHT SCENES

CIE XYZ values of each colored object and 
XnYnZn of reference white were measured 
under each light scene and normalized by the 
reference white to calculate L*a*b* coordinates 
using CIELAB equations [8].

L* a* b* C*
ab

h
ab

1.Fluo 2700K 40.74 -8.44 -32.79 33.86 1.32

2.Fluo 6500K 47.09 6.85 -73.70 74.02 -1.48

3.LED 2900K 34.53 -14.15 -65.54 67.05 1.35

4.LED 6100K 45.17 14.12 -60.83 62.45 -1.34

L* a* b* C*
ab

h
ab

1.Fluo 2700K 60.43 -10.58 -16.84 19.88 1.01

2.Fluo 6500K 63.75 -7.45 -60.21 60.67 1.45

3.LED 2900K 58.57 -17.14 -38.36 42.01 1.15

4.LED 6100K 64.14 -3.68 -34.89 35.08 1.47

L* a* b* C*
ab

h
ab

1.Fluo 2700K 51.82 -46.39 40.21 61.39 -0.71

2.Fluo 6500K 31.51 -61.32 24.41 66.00 -0.38

3.LED 2900K 46.18 -61.33 37.27 71.77 -0.55

4.LED 6100K 48.58 -64.74 49.76 81.65 -0.65

L* a* b* C*
ab

h
ab

1.Fluo 2700K 94.95 10.21 93.53 94.09 1.46

2.Fluo 6500K 83.97 -15.51 66.80 68.58 -1.34

3.LED 2900K 95.63 11.18 99.63 100.26 1.46

4.LED 6100K 90.14 -2.03 103.04 103.06 -1.55

L* a* b* C*
ab

h
ab

1.Fluo 2700K 90.49 50.73 76.57 91.85 0.98

2.Fluo 6500K 77.96 36.99 37.00 52.32 0.78

3.LED 2900K 91.74 49.75 87.60 100.74 1.05

4.LED 6100K 84.26 43.56 80.22 91.29 1.07

L* a* b* C*
ab

h
ab

1.Fluo 2700K 68.17 53.04 49.07 72.26 0.75

2.Fluo 6500K 56.54 40.84 15.46 43.66 0.36

3.LED 2900K 70.03 62.95 67.73 92.47 0.82

4.LED 6100K 58.95 58.80 51.42 77.11 0.72

Figure 17 – L*a*b*coordinates, 
chroma and hue values of blue. As 
a general rule, fluorescent lamps 
give less redness-greenness to the 
perception of the colored object. 
Brightness is higher with cold lights. 
It seems that Fluo 6500K is the best 
one because it maximizes all values, 
in contrast with previous results (LED 
6100K).

Figure 18 – L*a*b*coordinates, 
chroma and hue values of light blue. 
Brightness and hue are slightly higher 
with cold lights. Chroma is much more 
higher with Fluo 6500K.

Figure 19 – L*a*b*coordinates, 
chroma and hue values of green. In 
this case, every light scene maximizes 
a different value, so it’s not possible to 
define the best light, as it was noticed 
in the previous analysis.

Figure 20 – L*a*b*coordinates, 
chroma and hue values of yellow. 
LED 6100K seems to be the best 
light since all the values are at their 
maximum, except for brightness that 
is maximized with warm lights. The 
only scene completely different from 
the others is number 2.

Figure 21 – L*a*b*coordinates, 
chroma and hue values of orange. As 
the graphic shows, there is not a big 
difference between scenes number 
1, 3 and 4 (even if LED 2900K 
maximizes brightness and chroma), 
with the second scene different from 
the others once again.

Figure 22 – L*a*b*coordinates, 
chroma and hue values of red. 
Looking at the values and the graphic, 
red is nearer to orange that to the 
red-a* axe: in this case too, LED 
2900K maximizes brightness, a* 
value and chroma. Fluo 2700K and 
LED 6100K have good values too 
while Fluo 6500K is the worst light 
again.



50 Cultura e Scienza del Colore - Color Culture and Science | 03/15  

The next diagrams are CCT and technology 
comparisons among the 4 scenes, showing all 
color coordinates for each light.

In the CCT comparison, warm lights tend to move 
to the yellow-red area, even if it’s less evident for 
LEDs. In the second diagram, it’s possible to note 
that LED 2900K reduces the distance among 
yellow, orange and red coordinates.

In the technology comparison, the diagrams 
have the same shape, with a bigger chromatic 
variety for LEDs. For cold lights, LED diagram is 
also moved towards yellow-red area.

4. C0NCLUSIONS

From the obtained results it can be inferred that:

•  different lamps technologies with the 
same CCT produce different effects on 
colored objects;

•  under the same technology, but with 
different CCTs, different effects are 
obtained as well.

Color differences regard chroma, hue and value 
and, under the same conditions, depend on the 
spectral reflectance of the considered object.

This study shows that the only use of parameters 
like CRI and CCT is not enough to evaluate 
effects on colored objects. Moreover, further 
research is necessary to compare these results 
to experimental evaluations.

ACKNOWLEDGEMENT
The authors thank Zumtobel for providing 
luminaires and Luxmate control system.

BIBLIOGRAPHY 

[1] Boyce, P. R. (2003). Human factors in lighting (2nd ed.). 
London: Taylor & Francis.

[2] Cuttle, C. (2004). Brightness, lightness, and providing a 
preconceived appearance to the interior. Lighting Res. and 
Technol., 36(3), 201-216.

[3] Quality of the Visual Environment Committee. (2009). 
Light + design: A guide to designing quality lighting for 
people and buildings (DG-18). New York: Illuminating 
Engineering Society of North America.

[4] CEN (2011) - EN 12464-1- Light and Lighting - Light in 
work places - Part 1: indoor work places. Brussels.

[5] Boyce PR, Cuttle C.(1990) Effect of correlated colour 
temperature on the perception of interiors and colour 
discrimination. Lighting Res. Technol 22, 19-36.

[6] Fotios S.A. (2006) Chromatic adaptation and the 
relationship between lamp spectrum and brightness. 
Lighting Res. Technol. 38(1),  3-17.

[7] Navvab MA. (2001).Comparison of visual performance 
under high and low color temperature fluorescent lamps. 
Journal of the Illuminating Engineering Society, 30: 170–
175.

[8] Fairchild M.D. (2005) Color Appearance Model, (2nd 
ed.). John Wiley & Sons Ltd.

Figure 23 – CCT comparisons 
between Fluo 2700K and Fluo 6500K 
(A), LED 2900K and LED 6100K (B).

Figure 24 – Technology comparisons 
between Fluo 2700K and LED 2900K 
(A), Fluo 6500K and LED 6100K (B).