Colour Quality Is an Engineering Outcome, Not a Bulb Choice
Colour rendering is often treated as a single figure to be maximised. In practice it is the result of several independent decisions in the LED selection and optical design, and improving one without the others produces a fixture that scores well on a datasheet and looks wrong in a room.
The four properties below govern how a commercial fixture renders colour. They are measured differently, and they fail in different ways.
| Property | What it describes | Typical failure mode |
|---|---|---|
| CRI (Ra) | Average fidelity across eight to fourteen pastel test colours | Scores well while deep reds and saturated colours remain dull |
| R9 | Fidelity of saturated red specifically | Skin tones, wood and food appear flat or grey even at high Ra |
| TM-30 (Rf and Rg) | Fidelity plus gamut across a larger, more realistic sample set | Reveals saturation distortions that Ra averages away |
| CCT and SDCM | Warmth of the light, and how tightly production batches cluster | Visible inconsistency across a continuous ceiling plane |
Why R9 Deserves Separate Attention
CRI averages its test colours, and those colours are predominantly desaturated. A fixture can therefore reach Ra 85 while rendering saturated red poorly, because red contributes only part of the average. The commercial consequence is specific: retail displays of food, timber, leather and cosmetics are dominated by reds and warm browns, and those are exactly the materials that look dull when R9 is low. For mid-to-high-end retail and hospitality interiors, R9 above 50 is a reasonable floor, and R9 above 80 is what distinguishes a genuinely high-fidelity source.
Reading TM-30 Alongside CRI
TM-30 reports two values: Rf, which is fidelity, and Rg, which is relative gamut. Rf near 100 means colours shift little from a reference source. Rg above 100 means colours appear more saturated than the reference; below 100, less. Neither is inherently better — but both are informative, because a source with modest Rf and high Rg can flatter merchandise while a source with the same Rf and low Rg makes it look tired. Where a client has an existing reference installation, matching Rg as well as Rf is often what makes the new work read as consistent with the old.
Glare: The Comfort Problem That Brightness Cannot Solve
Glare complaints rarely mean the lighting is too dim. They usually mean the source is visible at an angle where the eye cannot tolerate it, and no amount of additional light elsewhere will correct that.
Three mechanisms are at work, and each needs a different remedy.
Direct Glare
Unshielded brightness in the field of view. Controlled by recess depth, baffles, louvres or lens design — that is, by geometry rather than by reducing output. A specification that addresses glare by lowering wattage trades comfort for illuminance and seldom satisfies either requirement.
Reflected Glare
Reflection of a source in a glossy surface: screens, polished stone, glass display cases. Controlled by the fixture's position relative to the viewing angle, which is a layout decision rather than a product one, and which is why a fixture's suitability cannot be assessed from its catalogue entry alone.
Discomfort and UGR
UGR is the systematic way to express the tolerance limit for a given interior. A worked scheme names the limit — commonly 19 for offices and 22 for circulation — and demonstrates compliance through the layout software rather than by assertion.
Uniformity and the Spacing Ratio
Uniformity is the ratio of minimum to average illuminance across the task area. Where a scheme specifies an average figure alone, a layout can satisfy it while leaving dark bands between fixtures, because bright pools raise the average without filling the gaps.
Two ratios govern the result:
- Spacing-to-height ratio — the maximum spacing at which adjacent fixtures still overlap. Published per fixture type, and the single most useful layout number.
- Minimum-to-average ratio — the design target, typically 0.6 or better for general lighting and higher where tasks are visually demanding.
A useful cross-check on a proposed layout is to compare the proposed spacing against the fixture's published spacing-to-height ratio. Where the proposal exceeds it, the average illuminance may still be met while the minimum is not.
Colour Temperature and the Application
| Application | Typical CCT | Reasoning |
|---|---|---|
| Hotel guest rooms and lounges | 2700–3000K | Warmth supports the intended sense of rest |
| Retail and display | 3000–3500K | Balances material rendition against perceived brightness |
| Open office and clinical | 3500–4000K | Higher CCT reads as more alert without increasing illuminance |
| Architectural and exterior | Selected to match the material | Warm light on stone and timber, neutral on glass and metal |
The consistency rule matters more than the specific choice: within a single visual field, mixing CCTs produces a difference the eye reads as a defect. Where a scheme requires two CCTs, they should be separated by a physical transition rather than placed adjacent.
The Dimming Behaviour Nobody Specifies
A fixture's colour behaviour under dimming is rarely published and frequently disappointing. Two issues arise:
- Colour shift — as drive current falls, some LEDs shift noticeably warmer or greener. On an evening scene dimmed to 10 %, the shift can be the most visible property of the installation.
- Flicker at low levels — flicker typically worsens as dim level falls, precisely at the levels used in hospitality and residential settings.
Both should be specified as measured values at the dim levels the design actually uses, not at full output.
Optics and the Character of the Light
Two fixtures with identical photometric outputs can produce very different impressions, because the distribution between beam and spill differs. A narrow beam with a sharp cut-off produces definition and shadow; a wide beam with a gradual fall-off produces evenness and a flatter rendering of surface texture.
- Beam angle — the angle at 50 % peak intensity. Defines the visual pool.
- Field angle — the angle at 10 %. Defines where the light effectively ends, and therefore the spacing at which pools merge.
- Cut-off quality — how sharply the beam terminates. A soft edge blends; a hard edge defines.
Building a Colour Specification That Holds
The specification below is short enough to enter a purchase document and specific enough to be verifiable.
- CRI (Ra) ≥ 90 with R9 ≥ 50, both measured per fixture
- CCT 3000K, 3 SDCM maximum, guaranteed per production batch
- TM-30 values reported so Rf and Rg can be compared against any reference installation
- UGR limit stated for the application, demonstrated by layout
- Minimum-to-average uniformity ≥ 0.6 across the task area
- Colour shift and flicker index stated at 10 % dim level
- Spacing-to-height ratio published for each fixture type used
Verifying After Installation
Specifications are only meaningful if they are checked. The measurements below are practical to take on a completed installation and directly test the properties discussed above.
- Illuminance at the working plane, sampled on a grid, to confirm average and minimum-to-average uniformity
- CCT and CRI spot-checked across fixtures from different production batches to confirm batch consistency
- Dimming sweep to the lowest used level, observing colour shift and flicker
- Glare assessment from the normal viewing positions, not from directly beneath a fixture
Summary
Colour quality is settled by three decisions made early: the LED bin and its R9 behaviour, the SDCM tolerance guaranteed across production, and the optical design that determines how the light reaches the surface. Each is measurable, each can be written into a specification, and each is cheaper to fix before tooling than after installation. For a mid-to-high-end commercial project, the colour specification is what distinguishes a lighting scheme that photographs well from one that continues to look correct after the fixtures have been running for a year.