UNEP-lites.

asia Laboratory Training Workshop

Beijing, China
22-24 April 2015
 UNEP GELC Lamp Performance Testing Training Workshop
April 22-24, 2015, Beijing

Fundamentals of Colorimetry and

Practical Color Measurements

Yoshi Ohno
CIE VP-Technical Elect, President-Elect
NIST Fellow, Sensor Science Division
National Institute of Standards and Technology
Gaithersburg, Maryland

UNEP GELC Training 2015 2

 Outline

1. Background of CIE 1931 colorimetry system

2. Chromaticity diagrams and MacAdam ellipses
3. CCT and Duv
4. ANSI chromaticity specification for SSL
5. Preferred white light chromaticity
6. Object color specification, CIELAB space
7. CRI and color quality
8. Practical color measurements of LED products

UNEP GELC Training 2015 3

 Outline

1. Background of CIE 1931 colorimetry system

2. Chromaticity diagrams and MacAdam ellipses
3. CCT and Duv
4. ANSI chromaticity specification for SSL
5. Preferred white light chromaticity
6. Object color specification, CIELAB space
7. CRI and color quality
8. Practical color measurements of LED products

UNEP GELC Training 2015 4

 Three cone sensitivities

UNEP GELC Training 2015 5

 Color Matching Experiments

Wright (1929) Guild (1931)

lR=650 nm Broadband
lG=530 nm Primaries
lB=460 nm (lamp+filters)

2°

UNEP GELC Training 2015 6

 Results

Wright (1929) Guild (1931)

• Monochromatic • red, green, blue filters
primaries (460, 530,
plus lamp
650 nm)
• 10 observers • 7 observers
• 2°field of view • 2°field of view

B B
G G R
R

UNEP GELC Training 2015 7

 CIE 1931 XYZ Color Matching Functions
(CIE 1931 Standard Colorimetric Observer)

2. 0

f (l )
()
x牋?
()
y牋?
Tristimulus Values
()
z牋?
X = k ò f (l )x (l )dl
1. 0 l

Y = k ò f (l ) y (l )dl
l

Z = k ò f (l ) z (l )dl
l
0. 0
350 400 450 500 550 600 650 700 750
W avelengt h ( nm )

2° field of view observer

(applicable to 1° to 4° field of view)

UNEP GELC Training 2015 8

 CIE 1964 10° Color Matching Functions
(CIE 1964 Supplementary Standard Colorimetric Observer)

Tristimulus Values

X10 = k ò f (l )x10 (l )dl

l

Y10 = k ò f (l ) y (l )dl
l 10

Z10 = k ò f (l ) z (l )dl
l 10

• Applicable to field of view greater than 4°.

• Used for some applications in object color, but not used
for light source specification.

UNEP GELC Training 2015 9

 Chromaticity coordinates

Tristimulus Values Chromaticity Coordinates

X
X x=
X +Y + Z
Y
Y
Z y=
X +Y + Z

•Y is a measure of visual intensity of light stimulus.

• x, y, Y fully describes a light stimulus.

UNEP GELC Training 2015 10

 Outline

1. Background of CIE 1931 colorimetry system

2. Chromaticity diagrams and MacAdam ellipses
3. CCT and Duv
4. ANSI chromaticity specification for SSL
5. Preferred white light chromaticity
6. Object color specification, CIELAB space
7. CRI and color quality
8. Practical color measurements of LED products

UNEP GELC Training 2015 11

 CIE 1931 (x, y) Chromaticity Diagram

Spectrum
locus
Color mixing

Chromaticity of a
mixture of two lights
lies along the line
between the two
chromaticity points
of the lights.
Purple
line

UNEP GELC Training 2015 12

 MacAdam Ellipses

Just noticeable
color differences.
(magnified by 10
times)

If the color space

were uniform,
these ellipses
would be circles
of the same size.

UNEP GELC Training 2015 13

 CIE 1960 (u, v) Chromaticity Diagram
(Now obsolete)
In 1960, the CIE
developed Uniform
Chromaticity Scale
diagram, based on the
work of MacAdam.

4X
u=
X +15Y + 3Z
6Y
v=
X +15Y + 3Z
4x
u=
(-2x +12y + 3)
6y
v=
(-2x +12y + 3)

UNEP GELC Training 2015 14

 CIE 1976 (u’, v’) Chromaticity Diagram

In 1976, CIE adopted an

amended Uniform
Chromaticity Scale which
gave better agreement
with experimental data:

u' = u ; v' =1.5v

4X
u¢ =
X +15Y + 3Z
9Y
v¢ =
X +15Y + 3Z
4x
u¢ =
(-2x +12y + 3)
9y
v¢ =
(-2x +12y + 3)

UNEP GELC Training 2015 15

 CIE (u’,v’) for chromaticity difference specification
CIE 1931 (x, y) Diagram CIE 1976 (u’, v’) Diagram

7 step
MacAdam 7 step
Ellipses MacAdam
Ellipses
R=0.008

7-step MacAdam ellipses »a circle with radius 0.008 on (u’v’) diagram.

Do not use MacAdam ellipses. x-step MacAdam ellises »radius 0.00x
UNEP GELC Training 2015 16
 CIE TN 001 (2014)

Primary author Y Ohno

Published July 2014

UNEP GELC Training 2015 17

 CIE TN 001 (2014)

The u'v' circle is specified with a centre point

and radius r on the (u',v') diagram, and expressed by,

Recommended to replace n-step

MacAdam ellipses

Available free at: http://files.cie.co.at/738_CIE_TN_001-2014.pdf

UNEP GELC Training 2015 18
 Conversions

(x,y)  (u’,v’) (x,y)  (u,v)

4x 4x
u¢ = u=
(-2x +12y + 3) (-2x +12y + 3)
9y 6y
v¢ = v=
(-2x +12y + 3) (-2x +12y + 3)

(u’,v’)  (x, y) (u,v)  (x, y)

9u¢ x=
9u
x=
(6u¢ -16 v¢ +12) (6u - 24v +12)
2v¢ y=
3v
y=
(3u¢ - 8v¢ + 6) (3u -12v + 6)

UNEP GELC Training 2015 19

 Outline

1. Background of CIE 1931 colorimetry system

2. Chromaticity diagrams and MacAdam ellipses
3. CCT and Duv
4. ANSI chromaticity specification for SSL
5. Preferred white light chromaticity
6. Object color specification, CIELAB space
7. CRI and color quality
8. Practical color measurements of LED products

UNEP GELC Training 2015 20

 Color Temperature
Temperature [K] of a Planckian radiator whose radiation
has the same chromaticity as that of a given stimulus.

Planckian radiation 1000 K

5.0
2000 K
Relative Spectral Power Distribution

3000 K
4.0
5000 K

10000 K
3.0
20000 K

2.0

1.0

0.0
300 400 500 600 700 800

Wav elength (nm)

UNEP GELC Training 2015 21

 Correlated Color Temperature (CCT)

Temperature [K] of a Planckian radiator whose chromaticity is closest to

that of a given stimulus on the CIE (u’,2/3 v’) coordinate.
(CIE 15:2004)

CIE (u’,2/3 v’) is the CIE

1960 (u, v) diagram,
which is now obsolete.

UNEP GELC Training 2015 22

 Chromaticity expression for lighting
CCT (Correlated Color Temperature)
Duv (Shift from Planckian locus)

Duv

(u’, v’) =
(0.245, 0.528)?

UNEP GELC Training 2015 23

 Duv
Duv scale on (u’, v’) diagram

Defined in ANSI C78.377

Closest distance from the Planckian locus on the (u', 2/3 v') diagram,
with + sign for above and - sign for below the Planckian locus.

UNEP GELC Training 2015 24

 CCT- Duv chart

4000 K 5000 K 6500 K

3000 K
2700 K 3500 K 4500 K 5700 K

7-step MacAdam
ellipses

(CCT in log scale)

UNEP GELC Training 2015 25

Direct approach (1) to calculate CCT and Duv
Triangular solution
(1) Create a table of CCT vs
distance di to BB locus on (u,v)
coodinate.
(2) Find the closest point in the
table.
Tx
Tm (3) Solve the triangle for the
neighboring 2 points
Tm+1
dm2 -1 - dm2 +1 + l 2
CCT u v distance d x=
2l
x
Tx = Tm -1 + (Tm +1 - Tm -1 ) ·
l

Tx Duv = [±sign] ( d 2
m -1 -x )
2 1/ 2

Use Planck’s equation and color matching functions at 1 nm interval.

UNEP GELC Training 2015 26

Direct approach (2) to calculate CCT and Duv
Parabolic solution
(1) Create a table of CCT vs
distance di to BB locus on (u,v)
coodinate.
(2) Find the closest point in the
Tx table.
Tm
(3) Parabolic fit for the neighboring
Tm+1
3 points.
d(T) = aT 2 + bT + C

(
Duv = [±sign] aTx 2 + bTx + C )

UNEP GELC Training 2015 27

 CCT Error in Triangular Solution

UNEP GELC Training 2015 28

 Conversion from (CCT, Duv) back to (x, y)

Input: CCT T (K) (u, v)

Duv Duv

1) Calculate (u0, v0) of the Planckian

radiator at T (K).
2) Calculate (u1, u1) of the Planckian (u0, v0)
radiator at T+DT (K). DT=0.01 (K)
3) Calculate
du = u1 - u0
(u1, v1)
dv = v1 - v0
u = u0 + Duv × sin q
u¢ = u
= u0 + Duv × dv / du + dv
2 2
v ¢ = 1.5v
v = v0 + Duv × cosq x = 9 u¢ /(6u¢ -16v ¢ + 12)
= u0 + Duv × du / du 2 + dv 2 y = 2 v ¢/(3u¢ - 8v ¢ + 6)
(Included in Revision draft of C78.377)
UNEP GELC Training 2015 29
 Simple calculation of Duv from (x, y) or (u’,v’)

Duv is normally calculated in the process of calculating CCT.

Below is a simple approximation formula, without calculation of CCT.

1) Convert (x, y) or (u’, v’) to (u, v)

u = 4 x/(-2x +12 y + 3) u = u¢
or
v = 6y/(-2x +12 y + 3) v = 2 v¢ / 3

2) Duv is obtained by

LFP = (u - 0.292) 2 + (v - 0.24) 2

æ u - 0.292 ö
a = arccosç ÷
è LFP ø
k6 -0.00616793
LBB = k 6 a + k5 a + k 4 a + k3 a + k2 a + k1 a + k0
6 5 4 3 2
k5 0.0893944
k4 -0.5179722
Duv = LFP - LBB k3 1.5317403
k2 -2.4243787
k1 1.925865
(Included in C78.377-2011) k0 -0.471106

UNEP GELC Training 2015 30

 Simple calculation from (x,y) or (u’,v’) to Duv

Accuracy of this method

within 0.00001 in the range

from 2600 K to 20000 K and
Duv 0.000 ± 0.010

within 0.0001 in the range

from 2160 K to 20000 K and
Duv 0.000 ± 0.010

(Included in C78.377-2011)

UNEP GELC Training 2015 31

 LEUKOS 10:1, 47-55, 2014 (DOI: 10.1080/15502724.2014.839020)

UNEP GELC Training 2015 32

 Outline

1. Background of CIE 1931 colorimetry system

2. Chromaticity diagrams and MacAdam ellipses
3. CCT and Duv
4. ANSI chromaticity specification for SSL
5. Preferred white light chromaticity
6. Object color specification, CIELAB space
7. CRI and color quality
8. Practical color measurements of LED products

UNEP GELC Training 2015 33

 ANSI and IEC (for Fluorescent Lamps)

ANSI C78.376-2001 IEC 60081 for Fluorescent Lamps

CIE 1931 (x, y) Diagram

UNEP GELC Training 2015 34

 ANSI C78.377 for Solid State Lighting Products
First published in 2008.
Used in Energy Star and many
regulations worldwide

2011 revision (current) 2015 revision (in ballot)

Duv
±0.006

UNEP GELC Training 2015 35

 4-step version in C78.377-2015 (Informative Annex)

Annex B. 4-step quadrangles Annex C. 4-step u’v’ circles

UNEP GELC Training 2015 36

 Outline

1. Background of CIE 1931 colorimetry system

2. Chromaticity diagrams and MacAdam ellipses
3. CCT and Duv
4. ANSI chromaticity specification for SSL
5. Preferred white light chromaticity
6. Object color specification, CIELAB space
7. CRI and color quality
8. Practical color measurements of LED products

UNEP GELC Training 2015 37

 Chromaticity below Planckian Locus
Current standard

Anecdotes say …
 Lights below Planckian
locus look better.

An example in neodymium
lamp

UNEP GELC Training 2015 38

 NIST Spectrally Tunable Lighting Facility

UNEP GELC Training 2015 39

UNEP GELC Training 2015 40
 2013 Vision Experiment at NIST
on Preferred and Acceptable level of Duv

Experiments made at 4 • 18 subjects participated.

CCTs, at 6 Duv points at • Subjects viewed fruits on
each CCT, at total 23 the table, his/her skin tone
points. and the whole room.
Total 50 spectra used. • Selected lights that looked
―more natural‖.
UNEP GELC Training 2015 41
 2013 Vision Experiment at NIST
on Preferred and Acceptable level of Duv

Results

Proposal made to ANSI

C78.377 to allow such products.

Another experiment planned for

summer 2015 at NIST.
Duv ≈
-0.015

Y. Ohno and M. Fein, Vision Experiment on Acceptable and Preferred White Light
Chromaticity for Lighting, CIE x029:2014, pp. 192 – 199 (2014)

UNEP GELC Training 2015 42

 Outline

1. Background of CIE 1931 colorimetry system

2. Chromaticity diagrams and MacAdam ellipses
3. CCT and Duv
4. ANSI chromaticity specification for SSL
5. Preferred white light chromaticity
6. Object color specification, CIELAB space
7. CRI and color quality
8. Practical color measurements of LED products

UNEP GELC Training 2015 43

 Object Color Measurement

Light source
S()
Relative power

Eye or detector

Wavelength (nm)

Reflected light
Reflectance factor
S(λ) • R(λ)

Relative reflection
R(λ)
Relative reflectance

Wavelength (nm)

Wavelength (nm)
sample

UNEP GELC Training 2015 44

 Tristimulus Values for Object Colors

Trisimulus values
X = k ò S(l ) R(l ) x(l )dl
l

Y = k ò S(l ) R(l ) y(l )dl

l

Z = k ò l S(l ) R(l ) z(l )dl

R(): Spectral reflectance factor of object surface
S(): Spectral distribution of illumination (standard illuminant).

ò l S(l ) y(l )dl

Y= 100 (%) for a
k =100 perfect diffuser.

Y gives luminance factor of the surface in %

(for the given illumination).

UNEP GELC Training 2015 45

 CIE Standard Illuminants
To Calculate Object Color, one of standard illuminants is used.
Standard Illuminant A:

Relative power
Representative of tungsten-filament lighting with
a color temperature of 2856 K.

Standard Illuminant D65: Wavelength (nm)

Representative of average daylight with a CCT

of ~6500 K.

Relative power
Other Daylight Illuminants D50, D55, D75

<Now obsolete> Wavelength (nm)

Illuminant B: direct sun light with a CCT of ~4900 K

Illuminant C: average daylight with a CCT of ~6800 K
( realized by a tungsten source with a prescribed liquid filter.)

*Formulae to calculate values of Illuminants A and D are available in CIE 15:2004

UNEP GELC Training 2015 46

 Light Color vs. Object Color


Chromaticity diagrams such as
(x,y), (u’,v’) are two-dimensional
and are only for light color. These
are not for object color.
No black, grey, or brown

Object color needs another axis: black—white

Object color needs a 3-dimensional color space.

UNEP GELC Training 2015 47

 Object Color Space

 Three attributes of object

color are hue, chroma

Lightness
(saturation), and lightness, white

and are expressed in a three

dimensional space.
Chroma

 To allow accurate Hue

specification of object colors
and color differences, CIE black

recommended
CIELAB and 3D color space
CIELUV
in 1976.

UNEP GELC Training 2015 48

 CIE 1976 (L*a* b*) color space
(CIELAB color space)
L*=100
X, Y, Z : for object surface
Xn , Yn , Zn : white reference
(perfect diffuser)
L* = 116 (Y / Yn )1/3 -16
a* = 500 éë(X / X n )1/3 - (Y / Yn )1/3 ùû
b* = 200 éë(Y / Yn )1/3 - (Z / Z n )1/3 ùû
(when X / X n , Y / Yn , Z / Z n > 0.008856)

L*=0
See CIE 15:2004 for more details.

UNEP GELC Training 2015 49

 Example: (a*,b*) plots of 1200 Munsell color samples

Ref. D65

UNEP GELC Training 2015 50

 Color difference formulae

CIELAB space

CIEDE2000 (Improved formula)

DE *00 See CIE 142:2001

• is occasionally used for displays

as they simulate object colors.
• is used for uncertainties of object
color measurements.
UNEP GELC Training 2015 51
 Comparison of object color spaces
Plot of 15 saturated CIE L*a* b* (CIELAB) or CIE L*u* v* (CIELUV)
Munsell samples are the current CIE recommendations.
(used in CQS).

Illum. A W*U*V* CIELAB CIELUV

(used in CRI, obsolete)

D65

UNEP GELC Training 2015 52

 Outline

1. Background of CIE 1931 colorimetry system

2. Chromaticity diagrams and MacAdam ellipses
3. CCT and Duv
4. ANSI chromaticity specification for SSL
5. Preferred white light chromaticity
6. Object color specification, CIELAB space
7. CRI and color quality
8. Practical color measurements of LED products

UNEP GELC Training 2015 53

 Investigating problems of the CRI
Color Rendering Index (CRI) CIE 13.3

Test source Reference source

CIE
Dxx
Same CCT [K]
Planckian Standard Daylight
(CCT<5000 K) (CCT > 5000 K)

#1 #2 #3 #4 #5 #6 #7 #8

Ra

#9 #10 #11 #12 #13 #14

R9

UNEP GELC Training 2015 54

 Problems of CRI
1. CRI (Ra) badly penalizes visually preferred lights

CCT 4929 K, Duv-0.001

CRI Ra = 70

2. Good CRI (Ra) score does not guarantee good color rendering

CCT 5020 K, Duv 0.000

CRI Ra = 82, R9 = -99

UNEP GELC Training 2015 55

 Color Quality Scale (CQS)
Proposed by NIST to improve CRI on these problems

 Improvement of CRI, produces one number score that

correlates well with perceived naturalness for real objects.
 15 saturated test color samples

 Update the old formulae in CRI

• CIELAB color space
• CMCCAT2000 chromatic adaptation
• 0 to 100 scale
• RMS averaging of color differences
 Saturation factor (address the Hunt effect)
Standards work for new metric is still on-going (in CIE, IES)
CQS is used as a tool for color quality design.
W. Davis and Y. Ohno, Color Quality Scale, Optical Engineering 49 (3), 033602 March 2010

UNEP GELC Training 2015 56

 CQS 9.0 EXCEL sheet (Color Rendering Simulation)

Used by many companies as a design tool for color quality

UNEP GELC Training 2015 57

 Looking at ―Luminous Efficacy of Radiation‖

B-Y + broad Red B-Y + narrow Red RGBA (simulation)

LER= 310 lm/W LER= 375 lm/W LER= 382 lm/W

~20 to 25 % increase
Narrowband theoretically more efficient
UNEP GELC Training 2015 58
 Outline

1. Background of CIE 1931 colorimetry system

2. Chromaticity diagrams and MacAdam ellipses
3. CCT and Duv
4. ANSI chromaticity specification for SSL
5. Preferred white light chromaticity
6. Object color specification, CIELAB space
7. CRI and color quality
8. Practical color measurements of LED products

UNEP GELC Training 2015 59

 Color Quantities of Light Sources

All light sources:

Chromaticity coordinates (x,y), (u’, v’)

White light sources:

Correlated color temperature Tc(K)
Duv Duv
Color Rendering Index (CRI) Ra

Narrow-band sources (LEDs):

Dominant wavelength d (nm)

UNEP GELC Training 2015 60

 Dominant Wavelength
Wavelength of the monochromatic stimulus that, when additively
mixed in suitable proportion with a specified achromatic stimulus,
yields a color match with the color stimulus considered.

Achromatic stimulus is usually equal energy spectrum:

(x,y)=0.3333, 0.3333)

then a line
If this is the connecting the two
colored and going to the
stimulus spectral locus ends
at the dominant
wavelength
and this is the
achromatic
stimulus,

UNEP GELC Training 2015 61

 Peak wavelength vs. Dominant wavelength

Dominant Peak wavelength

wavelength (628 nnm)
(618 nm)

UNEP GELC Training 2015 62

 Peak vs. Dominant Wavelength
Difference
Peak WL Dom. WL (peak-dom)
(nm) (nm) (nm)
Examples of real LEDs 443 450 -7
454 459 -5
457 462 -5
458 463 -6
463 468 -6
471 475 -3
476 479 -3
497 499 -2
504 506 -2
517 523 -6
522 529 -7
525 532 -7
530 538 -9
592 590 3
599 596 3
628 618 9
638 626 12

UNEP GELC Training 2015 63

 Conversion from Photometric to Radiometric Quantities
[cd] [W/sr]
[lm] [W]

Conversions can be made by knowing the relative

spectral power distribution S ) of the light
source:
Xv K m ò lS (l)V ( l)dl
=K = [lm/W]
ò l S(l)dl
Km = 683 [lm/W]
Xe

This ratio K is called luminous efficacy of radiation (LER).

Xv: photometric quantity (e.g., Iv [cd])

Xe: radiometric quantity (e.g., Ie [W/sr])

UNEP GELC Training 2015 64

 Measurement of Spatially-averaged color

A sphere-spectroradiometer directly measures the spatially

averaged color quantities.

< Sphere-spectroradiometer system >

UNEP GELC Training 2015 65

 Gonio-spectroradiometer

This method may be used when a sphere-

spectroradiometer system is not available, or the
test sample is too large for such a system.

1. A goniometer equipped with a

spectroradiometer
(gonio-spectroradiometer)

2. A goniometer equipped with a colorimeter

(gonio-colorimeter) …. This must be calibrated
against a spectroradiometer for each SSL product
measured.

LM-79

UNEP GELC Training 2015 66

 Calculation of spatially averaged color

Example for chromaticity x

LM-79

UNEP GELC Training 2015 67

 Sources of uncertainty in a spectrometer

• Uncertainty of reference standards (spectral irradiance,

total spectral radiant flux)
• Input optics geometry
• Wavelength scale error
• Bandpass and scanning interval
• Random noise
• Stray light
• Detector non-linearity
• Detector zero drift
Ref: Colorimetry – Understanding the CIE System, edited by J. Schanda, John
Wiley and Sons, pp.101-134 (2008).
Chapter 5 Spectral Color Measurement (Y. Ohno)
Appendix 2 Uncertainties in Spectral Color Measurement (G. Gardner)

UNEP GELC Training 2015 68

 Bandpass error
7.0 Cool White FL

Error in u’v’
Measured (5 nm BP)
6.0

5.0

4.0

3.0

2.0

1.0

0.0
350 450 550 650 750
Wavelength (nm)

1.2 LED model

Measured (10 nm BP)

1.0

0.8

0.6

0.4 • Bandwidth of 5 nm (FWHM) or less is

0.2 acceptable for colorimetry of most light sources.
0.0
• Error is proportional to the square of bandwidth
460 510
Wavelength (nm)
560
increase.

UNEP GELC Training 2015 69

 Bandpass error correction
Stearns and Stearns method (S-S method) (also, ASTM E308)
Corrected value: S0
1 12 120 12 1
S0 = M -2 - M -1 + M 0 - M1 + M 2
98 98 98 98 98

• Applies only to a triangular bandpass

• Bandwidth and scanning interval
must be matched.

UNEP GELC Training 2015 70

 Ohno-Gardner method for bandpass correction
Applicable to any bandpass functions, non-triangular, asymmetric, not
matched with scanning interval.
I0 = ò s( l, l0 ) dl

I1 = ò s( l, l0 )l dl

I2 = ò s( l, l0 ) l2 dl
æCalculated numerically ö
ç ÷
èfor any bandpass functionø

Corrected value S0 is obtained from the neighboring five points.

S0 = b-2 × M -2 + b-1 × M -1 + b0 × M 0 + b1 × M1 + b2 × M 2
a-1
2
a-1 a0 a1 a12
with b-2 = , b-1 = - , b0 = , b1 = - , b2 = ,
X X X X X
2
and X = a0 - 2a-1a1.
Y. Ohno, A Flexible Bandpass Correction Method for Spectrometers (AIC 2005)
J. Gardner, Bandwidth correction for LED chromaticity, Color Res. Appl. 31(5) 374-380

UNEP GELC Training 2015 71

 How to determine the bandpass of an array spectrometer

(1) Measure emission line. (3) Fit to a model

(2) Wavelength-reversed data. (4) Normalize it

brel (m, l0 )
b(m, l0 ) =
ò m brel (m, l0 )

Y. Ohno, Measurement of Bandpass for Array Spectrometers, Proc., CIE 27th Session, July 2011

UNEP GELC Training 2015 72

 Stray Light Error
Example
1.0E+01
RED # 9 9 5 6
Diode-array
1.0E+00
NI ST/ 5nm BP

1.0E-01
Relati v e power

x=0.7035(∆x= -0.0020)
1.0E-02 y=0.2951(∆y=+0.0006)
1.0E-03

1.0E-04
x=0.7055
1.0E-05 y=0.2945

1.0E-06
400 450 500 550 600 650 700 750
Wav elength (nm)
Correction methods
available

Ref: Y. Zong et al, Appl. Opt. 45-6 (2006)

UNEP GELC Training 2015 73

 THANK YOU for your
attention.

Contact: [email protected]

UNEP GELC Training 2015 74