In search of an improved method for assessing the colour rendering properties of light sources

Abstract

The original experimental work of CIE to establish a method for assessing the colour rendering properties of light sources has been analysed. The spectral power distribution of the Standard Warm White fluorescent lamp, used in that experiment to define a General Colour Rendering Index (Ra) equal to 50, is presented with extended spectral range and resolution. This lamp was used to compute new Ra indices when the CIE(1964)UVW formula was substituted by the CIELUV or the new, improved SVF colour space formula. Ten fluorescent and discharge lamps and two extra sets of colour samples were included in this investigation. The Special Colour Rendering Indices (Ri) for each sample of a set of 24 saturated colours show large deviations when the UVW formula is substituted by the SVF formula. Less variation is seen for the averaged Ra indices. However, for the Warm White de Luxe lamp investigated by the CIE committee, only the SVF formula yields a Ra index equal to that obtained by visual rating. The SVF formula includes a correction for chromatic adaptation that correlates well with physiological data. This improved adaptation transform makes it possible to use daylight illumination as a fixed reference source. This leads to drastic changes of the Ra-values, mainly due to a systematic degradation of Ra with decreasing correlated colour temperature. It is suggested that the use of the SVF formula, in combination with a set of 20 colour samples of constant lightness and saturation and with hues evenly distributed along the colour circle, will lead to an improved colour rendering index, tentatively called the Rb index. Our investigations indicate that, compared to the present CIE Ra index, only minor changes will result for the Standard, de Luxe and broad-band type of lamps. For modern energy saving lamps with irregular spectral power distribution, however, like three-three-band and most of the discharge lamps, the Rb index may deviate significantly and lead to more realistic values.

Get full access to this article

View all access options for this article.

References

CIE TC-3.2. Minutes from the 25th meeting, Paris, Nov. 1980, the 26th meeting, Berlin (Sept. 1981 ) and the 27th meeting, Oslo (Sept. 1982).

Pointer, M.R. , Color Res. & Appl., 6, 108 (1981).

Opstelten, J.J. , Lgt. Res. & Technol., 12, 186 (1980).

Mori, L. and Fuchida, T. , Color Res. & Appl., 7, 285 (1982).

Valberg, A. , Sallstrom, P. and Seim, T. , CIE Publ. No. 50, 218-223 (1980).

Seim, T. and Valberg, A. , Submitted to J. Opt. Soc. Am.

CIE Subcommittee report, Illum. Eng., 57, 471-491 (1962).

CIE Publication No. 13.2 (TC-3.2), (1974).

Wyszecki, G. and Stiles, W.S. , Color science (John Wiley & Sons, Inc., New York, 1967).

10.

MacAdam, D.L. , J. Opt. Soc. Am., 68, 121 (1978).

11.

Bodmann, H.W. , Haubner, P. and Marsden, A.M. , CIE Publ. No. 50, 99-102 (1980).

12.

Valberg, A. and Seim, T. , Biol. Cybern., 46, 149-158 (1983).

13.

Seim, T. In preparation.

14.

The spectral power distribution for a set of 29 fluorescent and 37 discharge lamps has been obtained from J. J. Opstelten, N.V. Philps Gloeilampenfabrieken, Lighting Division, Eindoven, The Netherlands.

15.

The spectral power distribution of 12 typical fluorescent lamps submitted by Y. Watari and L. Mori to the CIE TC-1.3 Subcommittee on Standard Sources (1982). The FS30 lamp in this report was used for smoothing of the Standard Warm White lamp spectrum.

16.

These samples were made in the Laboratory of Colour Metrics of K. Miescher at the University of Basle. Their spectral reflectance distributions were measured at the Swedish Institute of Paper Research , Stockholm.

17.

The reflectance spectra for the samples were obtained from Macbeth/Kollmorgen Corp., Little Britain Road, P.O. Box 950, Newburgh, N.Y., USA, 12550.

18.

Valberg, A. and Seim, T. , Proceedings from 'Licht '82', 53-66, Lugano, Schweizerische Lichttechnische Gesellschaft, Switzerland (June 1982).