Evaluating Color Rendering With TM-30

Dr. Michael Royer, PNNL

ENERGY STAR Webinar
March 31, 2016

1
 1. How do I know how the colors in a space
will appear?

2. Will a given appearance be liked (or

perceived as natural, saturated, etc.)?

2
 Tools
Average Fidelity
CRI Calculation Engine
Ra (CRI)
Outdated Color Science Specific Sample Fidelity
Limited Color Samples R9

High Level Average Values

Fidelity Index (Rf)
Gamut Index (Rg)

Graphical Representations
TM-30 Calculation Engine
Color Vector Graphic
Modern Color Science Color Distortion Graphic
New Color Samples
Detailed Values
Skin Fidelity (Rf,skin)
Fidelity by Hue (Rf,h#)
Chroma Shift by Hue (Rcs,h#)
Fidelity by Sample (Rf,CES#)

3
 TM-30 Method for Color Rendition

Color Fidelity

The accurate rendition

of color so that they
appear as they would
under familiar
(reference) illuminants

Fidelity Index (Rf)

(0-100)

4
 Positive Hue Shift

Constant Fidelity (CRI)

Decrease Increase
Saturation Saturation
CRI = 80 Perfect
CRI = Fidelity
80

Negative Hue Shift

(Also possible to change

lightness, not shown)
5
 TM-30 Method for Color Rendition

Color Fidelity Color Gamut

The accurate rendition The average level of

of color so that they saturation relative to
appear as they would familiar (reference)
under familiar illuminants.
(reference) illuminants

Gamut Index (Rg)

Fidelity Index (Rf) ~60-140 when Rf > 60
(0-100)

6
 140
• Evaluate tradeoffs between
fidelity and saturation. 130

Decreased Saturation Increased Saturation

• When disparate fidelity and 120
gamut measures are used
110

Gamut Index, Rg
together, the tradeoffs are
less apparent. 100 Reduced Fidelity Reference
Illuminant

• But average values don’t 90

tell the whole story…
80

70

60
50 60 70 80 90 100
Fidelity Index, Rf

7
 TM-30 Method for Color Rendition

Color Fidelity Color Gamut Gamut Shape

The accurate rendition The average level of Changes over

of color so that they saturation relative to different hues
appear as they would familiar (reference)
under familiar illuminants.
(reference) illuminants Color Vector Graphic,
Hue Bin Chroma Shift
Gamut Index (Rg)
Fidelity Index (Rf) ~60-140 when Rf > 60
(0-100)

8
 Rf = 75 | Rg = 100 | CCT = 3500 K Rf = 75 | Rg = 100 | CCT = 3500 K

Decreased
Increased
Saturation Saturation
Hue Shift

9
 Fidelity Index by Hue, Rcs,hj

Fidelity Index by Hue, Rcs,hj

100 100
80 80 79
85 83 83 85 82 83 87 84 81
78 76
80 74 73 74 74 73 80 74 75 72 75
72 70 72 72
68
72 71
63 64
60 60

40 40

20 20

0 0
40% 40%
Chroma Change by Hue, Rcs,hj

Chroma Change by Hue, Rcs,hj

30% 30%

20% 20% 16%

14%
11% 11% 12% 11%
9% 11% 10% 10% 9%
10% 5% 7% 10% 5%
4% 2% 3% 4% 2%
0% 0%
-2% -1%
-3% -3% -3% -4% -5%
-10% -5% -10% -6%
-8%
-11%
-13%
-20% -14% -20%

-30% -30%

-40% -40%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Same red fidelity, shift in opposite directions.

10
10
 CIE CRI (1965/1974) IES TM-30-15 (2015)

Fidelity Metric Only Fidelity, Gamut, Graphical, [1, 2, 3, 4, 5]

Detailed/Hues
[6, 7, 8, 9, 10]
CIE 1964 U*V*W* CAM02-UCS (CIE CAM02)

8 color samples 99 color samples [11, 12, 13, 14]

Medium chroma/lightness Uniform color space coverage

Spectral sensitivity varies Spectral sensitivity neutral
Munsell samples only Variety of real objects

Ref Illuminant Step Function Ref Illuminant Continuous

(Uses same reference sources, but blended
between 4500 K and 5500 K)

No lower limit for scores 0 to 100 scale (fidelity)

and inconsistent scales

11
 Which source is best?

12
 Experimental Room

13
 Experimental Room: Context

Lighting Conditions: 26
Illuminance: 20 fc
CCT: 3500 K (on Planckian)
Objects: Generic consumer goods, balanced hues
Application: Undefined
Participants: 19-65, 16 females 12 males
Rating Questions: Normal-Shifted, Saturated-Dull,
Like-Dislike

14
 Experimental Conditions

130

120

110
TM-30 Gamut Index, Rg

100

90

80

70
60 70 80 90 100
TM-30 Fidelity Index, Rf

15
 Experimental Conditions

130

120

110
TM-30 Gamut Index, Rg

100

90

80

70
60 70 80 90 100
TM-30 Fidelity Index, Rf

16
 We’re going to look at averages (means)….

…but the person to person differences are substantial!

Almost every source received ratings across the full range for each question.
(Normalness, Saturation, Preference)

If you’re a specifier, you get to decide what you like for the given space!

17
 Preference vs. Fidelity

8
Dislike R² = 0.06
7
Mean Preference Rating

2
Like
1
60 65 70 75 80 85 90 95 100
Fidelity Index (Rf)

18
 Preference vs. Fidelity/Gamut

130
Dislike Model r2 = 0.68

5.5 120
1
5.0 110

IES TM-30 Rg
p = 0.000
4.5 100 2 3

4.0 90

3.5 80

Like 70
60 70 80 90 100
IES TM-30 Rf
p = 0.042

19
 Gamut Shape/Red Rendering

Same Fidelity, Same Gamut, Significantly Different Rating.

20
 22

21
 21

22
 Preference for Increased Red Saturation…with limits.

8
Dislike R² = 0.81
7
Mean Preference Rating

2
Like
1
-30% -20% -10% 0% 10% 20% 30%
Hue Bin 16 Chroma Shift (Rcs,h16)

23
 Preference Model for this Experiment
7

Participant Preference Rating

Less Liked

3
More Liked R² = 0.9355
2
2 3 4 5 6 7
TM-30 Model Predicted Preference Rating

Best Model for Preference:

Like-Dislike = 7.396 - 0.0408(Rf) + 103.4(Rcs,h163) - 9.949(Rcs,h16)

24
 Summary

Context =

Normalness = Fidelity + Red Saturation

Rf > 80 0% < Rcs,h1 < 8% (Rf,h1 > 80)

Saturation = Red Saturation

Maximize Rcs,h16, Rcs,h1

Preference = Fidelity + Red Saturation

Rf > 74 0% < Rcs,h16 < 15% (Rg > 100)
25 0% < Rcs,h1 < 15%
 A Look at Existing Sources

140
Phosphor LED
130 Color Mixed LED
Hybrid LED
120 Standard Halogen
Experimental
Preferred Filtered Halogen
110 Zone
Triphosphor Fluorescent, 7XX
IES TM-30 Rg

100 Triphosphor Fluorescent, 8XX

Triphosphor Fluorescent, 9XX
90 Metal Halide

80

70

60
50 60 70 80 90 100
IES TM-30 Rf

26
 A Look at Existing Sources
50%

40%

30%

20%

10% Experimental
Preferred
Rcs,h16

Zone
0%

-10%

-20%

-30%

-40%

-50%

27
 Why so Few Red-Enhancing Sources?
1. Penalization by CRI

28
 Why so Few Red-Enhancing Sources?
1. Penalization by CRI
140
Rf, Rg (TM-30)
Ra, GAI (rescaled)
130

1
120

110
Gamut Index

100

90

80

70
40 50 60 70 80 90 100
Fidelity Index
29
 Why so Few Red-Enhancing Sources?
1. Penalization by CRI
140
Rf, Rg (TM-30)
Ra, GAI (rescaled)
130

120

110
Gamut Index

2
100

90

80

70
40 50 60 70 80 90 100
Fidelity Index
30
 Why so Few Red-Enhancing Sources?
1. Penalization by CRI
140
Rf, Rg (TM-30)
Ra, GAI (rescaled)
130

120

110
Gamut Index

100

90

80

70
40 50 60 70 80 90 100
Fidelity Index
31
 Why so Few Red-Enhancing Sources?
1. Penalization by CRI
CIE Ra
IES TM-30 Rf

32
 Why so Few Red-Enhancing Sources?
2. Efficiency Considerations

Ra 84, R9 = -7, LER 343 Ra 83, R9 = 21, LER 311

33
 Why so Few Red-Enhancing Sources?
2. Efficiency Considerations

34
 Common Commercially Available Sources
(Developed for CRI Ra):

F32T8/735 F32T8/835 Blue-Pump Phosphor LED (81 CRI)

Projected Rank: 22 of 26 Projected Rank: 16 of 26 Projected Rank: 18 of 26

Ra 74, LER 348 Ra 85, LER 343 Ra 83, LER 309

35
 Enhanced Sources
(Developed for CRI Ra and/or Gamut Area)
(Note different CCT)

LED (Patent Application) Neodymium Incandescent LED (Available Product)

 
Ra 80, LER 272 Ra 77, LER 136 Ra 87, LER 295

(Might be perfect for a

different application!)
36
 Case Studies

37
38
39
40
41
 Additional Resources
IES Technical Memorandum (TM) 30-15 (Includes Excel Calculators):
IES Method for Evaluating Light Source Color Rendition
http://bit.ly/1IWZxVu

Optics Express journal article that provides overview of the IES method:
Development of the IES method for evaluating the color rendition of light sources
http://bit.ly/1J32ftZ

Application webinar co-sponsored by US Department of Energy and Illuminating Engineering Society:

Understanding and Applying TM-30-15: IES Method for Evaluating Light Source Color Rendition
http://1.usa.gov/1YEkbBZ

Technical webinar co-sponsored by US Department of Energy and Illuminating Engineering Society:

A Technical Discussion of TM-30-15: Why and How it Advances Color Rendition Metrics
http://1.usa.gov/1Mn15LG

LEUKOS journal article supporting TM-30’s technical foundations:

Smet KAG, David A, Whitehead L. 2015. Why Color Space and Spectral Uniformity Are Essential for Color Rendering
Measures. LEUKOS. 12(1,2):39-50.
http://dx.doi.org/10.1080/15502724.2015.1091356

LEUKOS editorial discussing next steps:

Royer MP. 2015. IES TM-30-15 Is Approved—Now What? Moving Forward with New Color Rendition Measures. LEUKOS.
12(1,2):3-5.
http://dx.doi.org/10.1080/15502724.2015.1092752

Lighting Research and Technology, Open Letter:

Correspondence: In support of the IES method of evaluating light source colour rendition
(More than 30 authors)
http://dx.doi.org/10.1177/1477153515617392

DOE Fact Sheet on TM-30

http://energy.gov/eere/ssl/downloads/evaluating-color-rendition-using-ies-tm-30-15

DOE TM-30 FAQs Page:

http://energy.gov/eere/ssl/tm-30-frequently-asked-questions
42
 References
1. David A, Fini P, Houser K, Ohno Y, Royer M, Smet K, Wei M, Whitehead L. 2015. Development of the IES method for
evaluating the color rendition of light sources. Opt Expr 23(12):15888.
2. Rea MS, Freyssinier JP. 2010. Color Rendering: Beyond Pride and Prejudice. Color Research and Application 35(6):
401–409.
3. de Beer E, van der Burgt P, van Kemenade J. 2015. Another color rendering metric: do we really need it, can we live
without it? Leukos 12(1-2):51–59. DOI: 10.1080/15502724.2014.991793
4. Wei M, Houser KW, David A, Krames MR. 2016. Effect of Gamut Shape on Color Preference. CIE 2016 “Lighting
Quality & Energy Efficiency”, Melbourne, Australia, 2016.
5. Ohno Y, Fein M, Miller C. 2015. Vision experiment on chroma saturation for color quality preference. Proceedings of the
28th CIE Session. Manchester, UK. CIE Publication 216:2015 1(1).
6. Luo MR, Cui G, Li C. 2006. Uniform colour spaces based on CIECAM02 color appearance model. Color Research and
Application 31(4): 320–330.
7. Smet KAG, David A, Whitehead L. 2015. Why color space uniformity and sample set spectral uniformity are essential for
color rendering measures. Leukos 12(1–2):39–50.
8. Sándor N, Schanda J. 2006. Visual colour rendering based on colour difference evaluations. Lighting Res Technol
38(3):225–239. DOI: 10.1191/1365782806lrt168oa
9. Jost-Boissard S, Avouac P, Fontoynont M. 2015. Assessing the colour quality of LED sources: Naturalness,
attractiveness, colourfulness and colour difference. Lighting Res Technol 47:769–794. DOI:
10.1177/1477153514555882.
10. Luo MR, Gu HT, Liu XY, Liu HY, Wang BY. 2015. Testing colour rendering indices using visual data under different LED
sources. Proceedings of the 28th CIE Session. Manchester, UK. CIE Publication 216:2015 1(1).
11. David A. 2014. Color Fidelity of Light Sources Evaluated over Large Sets of Reflectance Samples. Leukos 10(2):59–75,
DOI: 10.1080/15502724.2013.844654
12. Žukauskas A, Vaicekauskas R, Ivanauskas F and others. 2009. Statistical approach to color quality of solid-state lamps.
IEEE Quantum Electronics 15(4):1189–1198.
13. van Der Burgt PJM, van Kemenade JTC. 2010. About color rendition of light sources: the balance between simplicity
and accuracy. Col Res App 35(2):85–93
14. Li C, Luo MR, Pointer MR, Green P. 2014. Comparison of real colour gamuts using a new reflectance database. Col Res
App 39(5):442-451.

43