Short answer
A spectroradiometer measures light itself — the spectral power emitted by a lamp, a display or the sky, in absolute radiometric units — rather than how a sample reflects a light the instrument supplies. It is the instrument for source colour: correlated colour temperature, chromaticity, colour rendering and luminance of lamps and screens. A spectrophotometer, by contrast, carries its own light and reports relative reflectance.
An object's colour depends on three things: the light falling on it, the object's reflectance and the observer. That is why a spectrophotometer brings a known light, measures reflectance as a ratio, and lets software apply any standard illuminant afterwards. A light source's colour depends on only two: its spectral power distribution and the observer. There is nothing to reflect and nothing to normalise to; the quantity of interest is the light's own spectrum, often in absolute units, with Y scaled to luminance using the maximum luminous efficacy of 683 lumens per watt. Measurement geometry for reflecting samples is irrelevant, but the viewing angle to the source can matter a great deal.
Lighting: a spectroradiometer's SPD is the input to correlated colour temperature, distance from the Planckian locus (Duv), the CIE colour rendering index and TM-30 fidelity and gamut metrics, all of which are calculated rather than measured directly. Displays: white point, primaries, luminance and tone response, and the reference readings used to correct cheaper colorimeters for a panel's backlight. Daylight and environments: sky and daylight spectra, which is how the CIE daylight series was originally derived. Some spectroradiometers are telescopic, measuring a small spot at a distance; others measure irradiance at a surface through a cosine-corrected diffuser.
Absolute measurement needs calibration against a standard lamp of known spectral output, traceable to a national metrology institute, and recalibration as the instrument ages. Narrow emission lines — in fluorescent and some LED lamps — are sensitive to the instrument's bandwidth and wavelength accuracy: a few nanometres of error shifts computed chromaticity. Stray light inside the instrument inflates readings in spectral regions where the source is weak. For displays, the measurement angle matters on panels whose colour changes off-axis, so the angle has to be fixed and reported.
For a quick CCT and illuminance check, a colour meter with filtered sensors is enough and much cheaper. For anything involving colour rendering, TM-30, light with spiky spectra, or a reference to correct other instruments, the spectrum is required and so is a spectroradiometer. For a surface colour — paint, fabric, print — neither is the right tool on its own: that needs a spectrophotometer, which supplies a defined light and geometry so that the result describes the material rather than whatever light happened to fall on it.
Why: Filtered colour meters approximate the colour-matching functions and err differently on spiky LED spectra.
Fix: Use a calibrated spectroradiometer as the reference and correct the meters to it.
Why: Panel warm-up, or measuring at a slightly different angle on a panel with angle-dependent colour.
Fix: Warm the display up, fix the measuring position and angle, and repeat readings.
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
FactStrong evidence
Measuring the colour of a light source requires only its spectral distribution and the observer's colour-matching functions, whereas object colour also requires the illuminant and the object's reflectance.
Source: Precise Color Communication; CIE 015:2018 Colorimetry, 4th edition
FactStrong evidence
When source colour is measured in absolute units, the tristimulus value Y is scaled to photometric quantities with a normalising constant of 683 lumens per watt.
Source: Precise Color Communication
FactStrong evidence
Colour rendering metrics such as the CIE colour rendering index and ANSI/IES TM-30 are computed from a light source's measured spectral power distribution.
Source: CIE 013.3-1995 Method of Measuring and Specifying Colour Rendering Properties of Light Sources; ANSI/IES TM-30-24, IES Method for Evaluating Light Source Color Rendition
FactStrong evidence
National metrology institutes such as NIST maintain the radiometric scales to which spectroradiometer calibrations are ultimately traced.
Source: NIST colorimetry and photometry
Reviewed 29 September 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.