Short answer
A colorimeter measures colour through three filters shaped like the eye's colour-matching functions and reports tristimulus values directly for one fixed light and observer. A spectrophotometer measures how much light a sample reflects (or transmits) at each wavelength and calculates colour from that spectrum for any illuminant and observer. Colorimeters are cheaper and fine for comparing like with like; spectrophotometers are needed for metamerism, formulation, and anything where the light or the material changes.
The CIE defines colour mathematically: the tristimulus values X, Y and Z are the light reaching the eye weighted by three colour-matching functions. A tristimulus colorimeter builds those weightings into hardware — a lamp, three (sometimes more) filtered detectors whose sensitivities approximate the functions — so its output is X, Y and Z for that built-in light, from which CIELAB follows. A spectrophotometer instead splits the reflected light with a grating or a bank of narrow-band sensors and records a reflectance factor every 10 nm or so across the visible range. The software then multiplies that curve by any illuminant and observer and integrates, arriving at X, Y and Z by calculation.
Two samples can produce identical tristimulus values under one light and differ under another — metamerism. A colorimeter, locked to one light, cannot see it coming; it reports a match. A spectrophotometer has the whole reflectance curve, so it can compute the colour under D65, D50, a tungsten lamp or a CIE LED illuminant and flag a metameric pair. The spectrum is also what colour-formulation software needs to predict a paint or dye recipe, and what print characterisation data are built from. A colorimeter's filters only approximate the colour-matching functions, so its absolute accuracy also varies with the spectral shape of the sample, especially for saturated or fluorescent colours.
When the job is to check that production parts match a standard of the same material, measured the same way under the same light, a colorimeter is often enough: filter errors largely cancel in a difference measurement between spectrally similar samples. Instrument makers position them for production-line and incoming-goods checks where portability, speed and price matter. Display calibrators are also colorimeters, though they measure emitted light rather than reflected, and they need corrections matched to each display's backlight spectrum for the same reason — a filter fitted to one spectrum is inaccurate on another.
Measuring the full spectrum does not make readings independent of how they were taken. Geometry (45°/0° or sphere), specular inclusion, aperture size, UV content of the lamp for fluorescent samples, and backing for thin samples all change the numbers, and two spectrophotometers of different design can disagree by a ΔE00 or more on the same sample. The spectrum is a better description of the material, but it is still a measurement made under stated conditions, and those conditions must be recorded and matched for results to be comparable.
| Instrument | Measures | Illuminant and observer | Detects metamerism | Typical use |
|---|---|---|---|---|
| Tristimulus colorimeter | Filtered X, Y, Z of reflected or emitted light | Fixed by hardware | No | Production checks; display calibration |
| Spectrophotometer | Spectral reflectance or transmittance of an object | Any, chosen in software | Yes | Formulation, print, QC, research |
| Spectroradiometer | Absolute spectral radiance or irradiance of a light | Not needed for sources; observer chosen | For light sources | Lamps, displays, daylight |
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
A tristimulus colorimeter measures light through sensors filtered to approximate the CIE colour-matching functions, while a spectrophotometer measures spectral reflectance at many wavelengths and computes tristimulus values from it.
Source: Precise Color Communication; CIE 015:2018 Colorimetry, 4th edition
FactStrong evidence
Colorimeters are not appropriate for analyses such as metamerism and colorant strength, which need spectral reflectance data.
Source: Precise Color Communication
FactStrong evidence
Tristimulus values for an object colour are calculated by integrating the product of the illuminant's spectral power distribution, the sample's spectral reflectance and the observer's colour-matching functions over wavelength.
Source: CIE 015:2018 Colorimetry, 4th edition; CIE datasets (colour-matching functions, illuminants)
Colourwise interpretationModerate evidence
For routine pass/fail checks of samples against a standard of the same material, a colorimeter's filter errors largely cancel, which is why it is adequate there but not for absolute or cross-material measurement.
Based on: Follows from the colorimeter's fixed approximation to the colour-matching functions: errors depend on the sample spectrum, so they are similar for spectrally similar samples and differ for dissimilar ones.
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.