What colour temperature actually describes
Colour temperature is the colour of the light compared with the colour of a theoretical black body heated to a temperature. A low value reads warm and slightly orange, which is the light most people associate with a campfire or an incandescent bulb. A high value reads cool and slightly blue, which is the light most people associate with daylight or an office. The number is a comparison, not a temperature of the lantern, and a buyer should read it as a description of hue rather than of heat.
There is a second property that is often reported alongside it and is easy to confuse with it: Duv, which describes how far the light sits from the black body curve. A light can sit close to the curve and still look a little green or a little pink to the eye, and the Duv figure is the number that carries that information. A specification that reports colour temperature and nothing else cannot tell the buyer whether the light will look warm and clean or warm and green.
What CRI adds
CRI is a different question. It measures how faithfully a light source reveals the colours of the objects under it, compared with a reference source of the same colour temperature. A high CRI light makes a red apple look red and a blue jacket look blue. A low CRI light makes both look muted, brownish or shifted, and the effect is most visible in the mid-saturation colours that people care about outdoors: skin, food, foliage, fabric, maps.
CRI is not determined by colour temperature. A warm light can render colour badly and a cool light can render it well, and the two numbers travel independently. That is the first boundary buyers need to hold in mind, because a specification that offers a pleasing warm colour temperature and omits CRI is offering a description of hue without saying whether the colours under it will survive.
For a tent, CRI matters more than the headline colour figure suggests. A camper reads a map, checks a trail marker, sorts food into a pan and looks at a companion's face, and all of those are colour tasks. A light that renders well makes those tasks easier and the tent feel more natural; a light that renders poorly makes the same tasks harder and the tent feel flat.
The four colour fields a specification should carry
| Field | What it describes | What it means if it is missing |
|---|---|---|
| Colour temperature | The nominal warmth of the light | The buyer cannot match the light to the setting |
| Duv or tint | How far the light sits from the black body curve | The light may read green or pink in use |
| CRI or Ra | How faithfully colours are revealed | Colour tasks may look muted or shifted |
| Bin or tolerance | The permitted spread across the array and the batch | A batch may visibly differ from the sample |
| Drive condition | The current at which the colour was measured | The colour shifts with drive and temperature |
| Sample identity | Golden sample or production batch | The colour may not describe what ships |
Ask for the fields, not the colour name. A supplier who can fill six rows has measured the colour and kept the record. A supplier who can only repeat "warm white" has not measured it, or has decided that the buyer does not need the number.
Binning and why the batch matters
LED dies are manufactured in bins. Each bin is a range of forward voltage, a range of output and a range of colour, and the bin is chosen at the array level before the lamp is assembled. A single model can therefore be built from a tightly binned batch or a loosely binned one, and the two will look different under the same nominal colour temperature.
Binning matters most at the step between batches. A first batch may be tightly binned because it was sorted carefully for a launch. A later batch may be loosely binned because the factory is running at volume and the sorting step was relaxed. The lantern still meets the nominal colour temperature on the specification sheet, but two lanterns bought six months apart may look visibly different side by side on a table.
The remedy is to write the bin into the specification. The nominal colour temperature is the target, the tolerance is the permitted spread, and the bin selection process is how the factory controls the spread between batches. A buyer who asks for the bin range and the tolerance gets a colour claim that survives a repeat order. A buyer who asks only for the colour temperature gets a claim that describes one batch. The drive current belongs in the same record, because the driver setting shifts the colour as well as the output.
Warm CCT ≠ good colour rendering
The boundary to state plainly: a warm colour temperature is not a sign of good colour rendering, and a cool colour temperature is not a sign of poor colour rendering. The two numbers describe different things. The first impression is about hue, the second is about fidelity, and conflating them is the single most common colour mistake on a specification sheet.
A second boundary: a nominal colour temperature is not a promise about every unit. Two lanterns in the same batch can differ slightly, and two batches bought at different times can differ more. The bin range and the tolerance are the numbers that describe the spread, and they are the numbers a buyer should hold the supplier to.
A third boundary: colour is not brightness. A light with a pleasing colour and a modest output is not the same purchase as a light with an impressive output and a poor colour. The lumen inflation and IP stacking article treats the brightness side of the same trade, and the mechanics of spec inflation treats why colour figures are sometimes presented without the conditions behind them.
Five questions for the supplier
- What is the nominal colour temperature, and at what drive current and junction temperature was it measured? Colour shifts with both.
- What is the Duv or tint figure, and what tolerance is permitted on the batch? The warmth alone does not describe the tint.
- What is the CRI or Ra figure, and how was it measured, including the reference source? A CRI figure without its method is a claim.
- What bin range is applied to the array, and how is bin consistency controlled between batches? This is the difference between one batch and every batch.
- Can you send the colour report and the bin declaration with the quotation rather than after the order? The document is the point of the exercise.
Next step
Request the six colour fields for every model under consideration, keep the report with the quotation, and repeat the same request at sample approval. When the colour arrives as a binned, tolerance-bounded figure, the comparison between two suppliers becomes a comparison of colour, not of names. Related reading: tent camping lighting for how colour combines with beam and runtime in a tent, the glossary of lumens and beam for the photometric vocabulary, and lumen inflation and IP stacking for the brightness side of the same trade. To apply this to a live sourcing project, start at contact.
Part of the Lumizuno outdoor-lighting notes. Illustrative values only; lumens and ingress ratings must be read with their test conditions.