Abstract

New and recently published data, comprising a total of approximately 30 000 judgments (ratio and paired comparisons) made by several observers on 176 pairs of colors, are used to check the 1964 CIE color-difference formula. The results indicate that the formula predicts observed color difference, ranging from about 4 to about 60 CIE units, with reasonable accuracy. Approximately 90% of the observed differences are predicted correctly to within ±25%. The precision of the observed differences is estimated to be also about ±25%. It has been noticed that several pairs (about 20 out of 100) have predicted color differences deviating significantly from the observed differences. Observations are affected by the luminous reflectance of the achromatic surround, but the deviations between color differences observed with a white surround and color differences observed with a black surround are masked somewhat by the limited precision of the observations, and thus, in the average, the CIE formula predicts both cases equally well.

© 1965 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. Commission Internationale de l’Eclairage, Proceedings of 15th Session, Vienna 1963 (Publication No CIE. 11 A, Bureau Central de la Commission, 57 rue Cuvier, Paris 5°, 1964), Vol. A, p. 35.
  2. This is a development from the 1960 CIE– UCS diagram Compt. Rend. A, 36 XIV° Session, Bruxelles, 1959).
  3. G. Wyszecki, J. Opt. Soc. Am. 53, 1318 (1963).
    [Crossref]
  4. Y. Sugiyama and H. Wright, J. Opt. Soc. Am. 53, 1214 (1963).
    [Crossref]
  5. Y. Sugiyama and H. Wright, J. Opt. Soc. Am. 54, 75 (1964).
    [Crossref] [PubMed]
  6. H. Wright, (To be published).
  7. Actually, Wright’s report discusses only seven of the nine groups.
  8. D. B. Judd and G. Wyszecki, Color in Business, Science, and Industry (John Wiley & Sons, Inc., New York, 1963), 2nd ed.

1964 (1)

1963 (2)

Judd, D. B.

D. B. Judd and G. Wyszecki, Color in Business, Science, and Industry (John Wiley & Sons, Inc., New York, 1963), 2nd ed.

Sugiyama, Y.

Wright, H.

Wyszecki, G.

G. Wyszecki, J. Opt. Soc. Am. 53, 1318 (1963).
[Crossref]

D. B. Judd and G. Wyszecki, Color in Business, Science, and Industry (John Wiley & Sons, Inc., New York, 1963), 2nd ed.

J. Opt. Soc. Am. (3)

Other (5)

H. Wright, (To be published).

Actually, Wright’s report discusses only seven of the nine groups.

D. B. Judd and G. Wyszecki, Color in Business, Science, and Industry (John Wiley & Sons, Inc., New York, 1963), 2nd ed.

Commission Internationale de l’Eclairage, Proceedings of 15th Session, Vienna 1963 (Publication No CIE. 11 A, Bureau Central de la Commission, 57 rue Cuvier, Paris 5°, 1964), Vol. A, p. 35.

This is a development from the 1960 CIE– UCS diagram Compt. Rend. A, 36 XIV° Session, Bruxelles, 1959).

Cited By

OSA participates in Crossref's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (10)

F. 1
F. 1

(u,υ) chromaticity points of six pairs of colors and source (S) used in Set 1 experiments.

F. 2
F. 2

Correlation between calculated and observed color differences (ΔEcalEobs) for the six pairs of colors of Set 1. Solid dots refer to paired comparison experiment, open circles to ratio comparison experiment. Points falling on the 45° line indicate perfect agreement; points falling between upper and lower boundary lines indicate agreement within ±25%.

F. 3
F. 3

(u,υ) chromaticity points of seven colors forming 21 pairs used in the Set 2 experiment.

F. 4
F. 4

Correlation between calculated and observed color differences (ΔEcalEobs) for the 21 color pairs of Set 2. Large dots refer to the mean of 10 trials; smaller dots refer to a single trial.

F. 5
F. 5

(u,υ) chromaticity points of 19 colors used in Set 3 experiment.

F. 6
F. 6

Correlation between calculated and observed color differences (ΔEcalEobs) for the 117 color pairs of Set 3.

F. 7
F. 7

(u,υ) chromaticity points of 32 pairs of colors used in Set 4 and 5 experiments. For each pair the color of higher luminous reflectance is indicated by an open circle, the other color by a solid dot.

F. 8
F. 8

Correlation between calculated and observed color differences (ΔEcalEobs) for the 32 color pairs of Set 4 (white surround).

F. 9
F. 9

Correlation between calculated and observed color differences (ΔEcalEobs) for the 32 color pairs of Set 5 (black surround).

F. 10
F. 10

Effect of changing the surround from white to black on the size of observed color differences as a function of the mean luminous reflectance Ym of a given color pair. The percent change is plotted as the ordinate, where ΔEW is the difference observed with a white surround, ΔEB the difference with a black surround.

Tables (9)

Tables Icon

Table I CIE coordinates of the six color pairs of Set 1.

Tables Icon

Table II Comparison of calculated color differences ΔEcal with observed color differences ΔEobs (Set 1).

Tables Icon

Table III CIE coordinates of the seven colors of Set 2.

Tables Icon

Table IV Comparison of calculated color differences ΔEcal with mean observed color differences ΔEobs (Set 2).

Tables Icon

Table V CIE coordinates of the 19 colors of Set 3.

Tables Icon

Table VI Comparison of calculated color differences ΔEcal with mean observed color differences ΔEobs (Set 3).

Tables Icon

Table VII CIE coordinates of the 32 color pairs of Sets 4 and 5.

Tables Icon

Table VIII Comparison of calculated color differences ΔEcal with observed color differences ΔEobs for 10 individual observers, two repeats, and the group (Set 4, white surround).

Tables Icon

Table IX Comparison of calculated color differences ΔEcal with observed color differences ΔEobs for 10 individual observers and the group (Set 5, black surround).

Equations (5)

Equations on this page are rendered with MathJax. Learn more.

W = 25 Y 1 3 17 ( 1 Y 100 ) U = 13 W ( u u 0 ) V = 13 W ( υ υ 0 )
u = 4 X / ( X + 15 Y + 3 Z ) ; υ = 6 Y / ( X + 15 Y + 3 Z ) ,
Δ E = [ ( U 1 U 2 ) 2 + ( V 1 V 2 ) 2 + ( W 1 W 2 ) 2 ] 1 2 .
α = i = 1 n t i ( Δ E i ) cal / i = 1 n t 1 2 .
σ = { 1 n 1 i 1 n [ ( Δ E i ) obs ( Δ E i ) cal ] 2 } 1 2 ,