Colour names, photographs and digital swatches can help describe a paint, but they are not objective measurements of the physical product.
The Chroma Institute starts with the paint itself. Physical paint samples are measured under controlled conditions and those measurements are converted into numerical colour data. This allows paints from different manufacturers and ranges to be analysed and compared on a common basis.
The result is a growing dataset designed around the measured colour of the physical paint itself.
Every measurement begins with a physical paint sample.
Rather than relying solely on manufacturer-provided colour values, photographs or colours displayed on a screen, The Chroma Institute measures prepared paint samples using instrumental colour measurement.
Consistent sample preparation and measurement procedures allow paints to be compared under the same conditions. This is particularly important when comparing products from different manufacturers, whose digital colour charts may have been produced using different methods.
The objective is simple: Measure every paint on the same basis.
Physical paint samples are prepared and measured under controlled conditions.
Instrumental measurements capture objective colour information from the physical sample. These measurements can then be transformed into standard colour coordinates for analysis.
Those measurements can then be analysed and compared across paints, ranges and manufacturers.
A colour cannot be adequately described by a name such as red, violet or olive green if we want to compare it objectively.
To analyse colours mathematically, they need to be represented numerically.
Paint Lab uses OKLab, a perceptual colour space in which each colour is represented using three coordinates.
L describes how light or dark a colour is.
Values progress from black towards white.
The a axis describes the colour's position between green and red.
Negative values extend towards green, while positive values extend towards red.
The b axis describes the colour's position between blue and yellow.
Negative values extend towards blue, while positive values extend towards yellow.
Together, these three coordinates give every measured paint a position within a three-dimensional colour space.
Digital colours are commonly communicated using RGB or Hex values.
These systems are extremely useful for specifying colours for digital display. A Hex colour is essentially a compact way of encoding RGB channel values.
But RGB and Hex were not designed to provide perceptually uniform coordinates for comparing how different two colours appear.
A Hex value does not directly tell us how light a colour appears, its perceptual hue or chroma, or how far it appears from another colour.
That distinction matters when colour becomes something we want to measure, analyse and compare, rather than simply display.
Designed primarily for: representing colours through red, green and blue components.
Excellent for: specifying and reproducing digital colours within a defined RGB colour space.
Limitation for Paint Lab: numerical distances between RGB values do not reliably correspond to perceived differences between colours.
Designed for: perceptual representation and manipulation of colour.
Excellent for: analysing colour relationships, changing lightness, chroma or hue, and calculating perceptual colour differences.
Advantage for Paint Lab: colours occupy positions within an approximately perceptually uniform space, allowing numerical relationships to correspond much more closely to visual relationships.
Put simply:
RGB and Hex provide ways of representing colour for digital display. OKLab provides Paint Lab with a perceptual coordinate system for analysing how measured colours relate to one another.
OKLab allows colours to be treated as positions within a three-dimensional map.
Lightness forms one dimension, while the green–red and blue–yellow axes provide the other two.
A measured paint therefore becomes more than a product name or digital swatch. It has a numerical position relative to every other measured paint.
Colours occupying nearby positions are perceptually similar. Colours occupying increasingly distant positions are increasingly different.
This spatial relationship forms the foundation of many of Paint Lab's colour-comparison tools.
OKLab can also be expressed using OKLCH coordinates.
Where OKLab uses:
L - Lightness
a - green ↔ red position
b - blue ↔ yellow position
OKLCH describes the same underlying colour using:
L - Lightness
C - Chroma
H - Hue
Chroma describes how far a colour lies from the neutral lightness axis, while hue describes its angular position around that axis.
This makes OKLCH particularly intuitive when thinking about familiar painting concepts such as lightness, chroma and hue, while OKLab provides convenient Cartesian coordinates for calculating distances between colours.
They are therefore two different coordinate systems for describing the same underlying colour space.
The familiar sRGB colour space used by traditional web Hex and RGB colours represents a defined range, or gamut, of reproducible colours.
But sRGB is not the entirety of visible colour, nor is it the only RGB colour space. Modern wide-gamut systems such as Display P3 can represent colours outside the traditional sRGB gamut. A wider display gamut can reproduce a greater range of colours, but it still does not make a screen equivalent to a physical painted surface.
OKLab provides a device-independent framework for representing and analysing colours without defining them simply as combinations of the red, green and blue channels of a particular RGB space.
Within Paint Lab, the measured colour remains the authoritative colour data.
When that colour needs to be shown on a screen, a digital representation is generated from the measured data.
Where the measured colour falls within the gamut of the display colour space, it can be represented accordingly. Where it falls outside that gamut, the display cannot reproduce the measured colour exactly.
This distinction is important:
The measurement is the data. The colour displayed on a screen is a representation of that data.
For this reason, Paint Lab's colour comparisons and calculations are based on measured colour data, not on the RGB or Hex values used to render colour swatches on screen.
The diagram below shows the approximate relationship between visible chromaticities and two common display gamuts. Select sRGB or Display P3 to see how each occupies only part of the wider CIE 1931 chromaticity space.
For decades, RGB and Hex have been the familiar language of colour on the web.
Modern colour technologies increasingly support perceptual and wide-gamut approaches alongside traditional sRGB. Modern CSS supports OKLab and OKLCH directly, bringing perceptual colour spaces into contemporary web technology.
These models are particularly useful when colours need to be manipulated, interpolated or compared in ways that behave more consistently with human perception.
Paint Lab applies the same principle to miniature paint.
Rather than treating a paint simply as a name attached to a Hex swatch, Paint Lab treats it as a measured position within perceptual colour space.
That measured position, rather than the screen swatch, becomes the foundation for comparison. The measurement tells Paint Lab what the colour is. The digital swatch allows the user to see a representation of it.
Once colours occupy positions within the same perceptual colour space, the distance between them can be calculated. Paint Lab uses ΔEOK to express that difference.
For two OKLab colours, the distance can be calculated from the differences between their L, a and b coordinates:
ΔEOK = √[(ΔL)² + (Δa)² + (Δb)²]
A smaller ΔEOK indicates that two measured colours occupy closer positions within OKLab. A larger value indicates greater separation. This gives Paint Lab a mathematical basis for searching the measured dataset and ranking paints according to their colour proximity.
Crucially, this calculation is performed using the underlying colour data, not by comparing the Hex or RGB values used to display the paints on screen.
Instead of asking:
"Which paints look similar in a digital swatch?”
Paint Lab can ask:
“Which measured paint is objectively closest according to ΔEOK?”
Paint ranges are not standardised. Different manufacturers may use similar names for noticeably different colours, while paints with very different names may occupy remarkably similar positions in colour space. By comparing measured colour coordinates rather than names or digital swatches, Paint Lab can identify related products across manufacturers and ranges. This makes it possible to investigate questions such as:
What is the closest measured alternative to this paint?
Which paints occupy this region of colour space?
How similar are these two paints?
Which measured paint is closest to the colour I need?
The comparison begins with measured colour, rather than product naming or screen appearance.
Digital colour is invaluable for displaying and communicating colour, but a digital representation is not itself a measurement of the physical paint. Photographs can be affected by lighting, camera settings, processing and white balance. Colours displayed digitally are affected by the source data, the colour capabilities of the display, its calibration and settings, and the conditions under which it is viewed.
Manufacturer colour charts remain useful product references, but different manufacturers do not necessarily produce them using a common measurement and display methodology. The Chroma Institute therefore starts with the physical product.
By preparing and measuring physical paint samples using a consistent methodology, the Institute can develop a dataset in which paints from different manufacturers can be analysed on the same basis.
There is an important distinction between measuring a paint and showing that paint on a screen. The measurement provides the underlying numerical colour data used by Paint Lab. A digital swatch is then generated to provide a useful visual representation of that measurement. The appearance of that swatch can vary between displays, and some measured colours may fall outside the range of colours a particular display can reproduce.
For this reason: Paint Lab does not use the appearance of the on-screen swatch to determine whether two paints match. It uses the underlying measured colour data.
The swatch helps the painter see the colour.
The measurement allows Paint Lab to analyse it.
The Chroma Institute is developing a standardised methodology for the preparation, instrumental measurement and analysis of miniature-paint samples. The methodology is designed to ensure that paints from different manufacturers and ranges can be measured consistently and compared on the same basis.
Measurements provide the underlying colour information retained within the Paint Lab dataset. From these data, perceptual colour coordinates can be calculated for analysis while appropriate digital representations can be generated for display.
Detailed information concerning sample preparation, measurement conditions, instrumentation, calibration, measurement geometry and data processing will be published as the measurement programme develops.
Our aim is to make the basis of the Paint Lab dataset clear, consistent and reproducible.
The usefulness of Paint Lab increases as more physical paints are measured.
The Chroma Institute is building its dataset across manufacturers, ranges, paint types and colours, creating an increasingly detailed map of the miniature-paint market.
Manufacturers can contribute physical product samples through the Manufacturer Product Sample Programme, allowing those products to be independently measured, classified and incorporated into the Paint Lab dataset.
Measurement is only useful if the resulting information can be put to practical use.
Paint Lab turns The Chroma Institute's measured colour dataset into tools that help miniature painters search for paints, compare products, find close colour matches, investigate colour relationships and build palettes.
The science sits underneath the application. Painters do not need to understand every coordinate or calculation to benefit from it.
They can simply make better-informed colour choices.