Correlated color temperature
Correlated color temperature (CCT) is the temperature of the Planckian radiator whose perceived chromaticity most closely corresponds to that of a specified light source under a defined colorimetric metric. It is expressed in kelvins, although the source itself need not be thermal and need not possess the spectral distribution of a black body.
CCT reduces a multidimensional spectral power distribution to a single coordinate associated with the Planckian locus. Sources having the same CCT can therefore differ substantially in spectral composition, color rendering, and distance from the locus. The quantity describes a chromatic relationship rather than the physical temperature of an emitting object.
Physical and colorimetric basis
The spectral radiance of an ideal black body at thermodynamic temperature (T) is given by Planck's law:
[ L_{\lambda}(\lambda,T)
\frac{c_1}{\lambda^5} \frac{1}{\exp\left(c_2/\lambda T\right)-1}, ]
where (\lambda) is wavelength and (c_1) and (c_2) are the first and second radiation constants. Integration of this spectrum against the CIE 1931 color-matching functions produces the tristimulus values
[ X(T)=\int L_{\lambda}(\lambda,T),\overline{x}(\lambda),d\lambda, ]
[ Y(T)=\int L_{\lambda}(\lambda,T),\overline{y}(\lambda),d\lambda, ]
[ Z(T)=\int L_{\lambda}(\lambda,T),\overline{z}(\lambda),d\lambda. ]
Normalization gives the chromaticity coordinates
[ x(T)=\frac{X(T)}{X(T)+Y(T)+Z(T)}, \qquad y(T)=\frac{Y(T)}{X(T)+Y(T)+Z(T)}. ]
As (T) varies, these coordinates trace the Planckian locus in a chromaticity diagram. Low-temperature sections of the locus occupy the yellow-to-red region, while higher-temperature sections progress through approximately neutral chromaticities toward blue. This colorimetric ordering accounts for the inverse relation between CCT and the conventional visual descriptions of “warm” and “cool” lighting: lower CCTs are commonly described as warm, whereas higher CCTs are commonly described as cool.
A non-Planckian source does not generally lie on the locus. Its CCT is associated with a nearby locus point according to lines of constant correlated temperature, conventionally called isotemperature lines. The result depends on the color space and distance definition because geometric proximity is not invariant under transformations between chromaticity diagrams.
Historical development
The physical foundation of color temperature emerged from nineteenth- and early twentieth-century studies of thermal radiation. Wilhelm Wien established the displacement relation connecting black-body temperature with the wavelength of maximum spectral emission, while Max Planck derived the complete spectral distribution that now bears his name. Their work defined the temperature-dependent family of spectra from which the Planckian locus is obtained.
The extension from color temperature to correlated color temperature followed the development of quantitative colorimetry. Deane B. Judd analyzed chromaticity differences and the orientation of constant-temperature relations, while David MacAdam investigated perceptual nonuniformity in chromaticity space. These contributions established that visually meaningful proximity to the Planckian locus could not be represented adequately by an arbitrary Euclidean distance in the original (x,y) diagram.
During the 1960s, A. R. Robertson and You Watanabe developed a tabular computation based on the CIE 1960 Uniform Chromaticity Scale. Their formulation represented temperature by reciprocal temperature and located a source between adjacent isotemperature lines. The resulting interpolation scheme became known as the Robertson method and supplied a stable numerical definition over a broad temperature interval.
Later analytical approximations reproduced portions of the same relationship through polynomial or rational functions of chromaticity coordinates. Such approximations differ in their valid temperature ranges and in their treatment of chromaticities remote from the Planckian locus. Contemporary colorimetric standards instead define CCT together with the coordinate system, reference observer, and numerical procedure required for reproducibility.
Reciprocal temperature and interpolation
The curvature of the Planckian locus makes direct interpolation in kelvins inconvenient. Reciprocal temperature provides a more nearly uniform parameter:
[ r=\frac{10^6}{T}, ]
where (r) is expressed in reciprocal megakelvins, historically called mireds. A difference of 100 K has little chromatic significance at high temperatures but can represent a substantial displacement at low temperatures. Reciprocal temperature more closely follows the visual spacing of locus chromaticities across these regions.
In the Robertson formulation, reference points on the locus are tabulated in (u,v) coordinates at selected reciprocal temperatures. Each reference point has an associated isotemperature-line slope. For a measured chromaticity ((u,v)), signed distances from neighboring lines identify the interval containing the perpendicular intersection with the locus. Interpolation within that interval yields reciprocal CCT, which is then converted to kelvins.
The method treats the locus locally rather than attempting a single global algebraic fit. Its accuracy consequently depends on the density of the reference table, the numerical interpolation rule, and the color-matching functions used to generate the locus.
Distance from the Planckian locus
CCT alone does not state whether a chromaticity lies directly on the Planckian locus. The accompanying quantity (D_{uv}) represents signed distance from the locus in the CIE 1960 (u,v) diagram. Positive values lie on the side conventionally associated with a green displacement, while negative values lie toward a magenta displacement.
Two sources can share a CCT while having different (D_{uv}) values because the relevant isotemperature line extends on both sides of the locus. A chromaticity sufficiently remote from the locus can still possess a mathematical nearest point, but the resulting CCT no longer provides a compact perceptual description. Standards therefore associate CCT with specified distance limits or validity regions rather than treating every chromaticity as an equally representative correlated temperature.
The sign and numerical magnitude of a locus offset depend on the adopted chromaticity system. The (D_{uv}) notation specifically refers to the CIE 1960 Uniform Chromaticity Scale and is not interchangeable with a raw distance measured in CIE (x,y) coordinates or in the later CIE 1976 (u^\prime,v^\prime) space.
Interpretation of source spectra
A thermal source close to ideal black-body behavior has a physical temperature that can approximate its color temperature. An incandescent filament provides a common example because its emission is dominated by a continuous thermal spectrum, although filament emissivity and enclosure effects prevent exact identity with an ideal radiator.
Many other sources obtain their CCT without having a corresponding thermodynamic temperature. A fluorescent lamp combines line emission from a discharge with broadband phosphor emission. A phosphor-converted light-emitting diode combines a narrow semiconductor component with wavelength-converted radiation. Both sources can have chromaticities near the Planckian locus even when their spectral distributions differ sharply from thermal radiation.
Daylight forms a separate chromaticity sequence produced by direct solar radiation, atmospheric scattering, and cloud-dependent redistribution. At higher correlated temperatures, the daylight locus does not coincide exactly with the Planckian locus. The CIE standard illuminant D65, which represents a daylight phase with a nominal CCT near 6500 K, therefore has a spectrum and chromaticity defined independently of a 6500 K black body.
Relation to appearance and color rendering
CCT characterizes the chromaticity of the illuminating source rather than the colors of objects illuminated by it. Sources with identical CCT and (D_{uv}) can produce different object-color appearances when their spectral power distributions interact differently with surface reflectance functions. This distinction underlies separate measures such as the color rendering index and IES TM-30.
Perceived whiteness also depends on chromatic adaptation, luminance level, visual surroundings, and viewing duration. A source that appears yellowish in one adaptation state can serve as a perceptual white after adaptation, without any change in its CCT. Consequently, correlated color temperature is a source coordinate within a colorimetric model rather than a complete model of visual appearance.
The same limitation applies to imaging systems. A camera white-balance setting labeled by CCT encodes an assumed illuminant chromaticity, but the transformation of recorded sensor values additionally depends on spectral sensitivities and the selected color-rendering model. A single temperature coordinate cannot reconstruct an arbitrary illuminant spectrum from its chromaticity.
Metrological status
A reproducible CCT value requires a specified standard observer, a defined chromaticity space, and an explicit computational method. Changes to the color-matching functions alter the integrated tristimulus values, while changes to the geometric metric alter the point considered nearest to the Planckian locus. Numerically identical temperature labels obtained under different definitions are therefore not necessarily metrologically equivalent.
CCT is most informative when the source chromaticity lies near the Planckian locus and when its locus offset is reported separately. Within that domain, it supplies a compact coordinate for comparing nominally white light sources. It does not determine spectral composition, luminous output, rendering behavior, or physical emitter temperature.