Cri Ra is one number built from eight pastel samples. It answers one question, whether colors look roughly like they would under a reference source, and it stays near 80 for sources that fail badly on deep red. TM-30 keeps that question and adds another: how saturated the colors look. Both numbers come from 99 samples, and both are shipped with a graphic that shows where the error lands.
Two Questions, Two Metrics
Fidelity, written Rf, is the average match to a reference source across 99 color evaluation samples. A source identical to the reference scores 100. Gamut, written Rg, compares the average chroma of the test source with the reference. Rg of 100 means matched saturation, below 100 means colors are muted, above 100 means they are pushed. A source can score Rf 92 and Rg 105: accurate on average, but louder in some hues.
Why Eight Samples Were Never Enough
The CRI sample set leans toward unsaturated pastels. A phosphor mix that drops energy near 650nm can still hold Ra above 80, because only one of the eight samples, R9, reaches deep red. R9 is reported separately for that reason, but the hue range between red and orange is covered thinly. TM-30 samples spread across the full hue circle and reach far higher chroma, so gaps that CRI averages away show up as a low skin-tone or red-region score.
| Property | CRI Ra | TM-30 Rf | TM-30 Rg |
|---|---|---|---|
| Sample count | 8 | 99 | 99 |
| Reference | Incandescent-like | Daylight, CCT dependent | Daylight, CCT dependent |
| Measures | Fidelity | Fidelity | Chroma change |
| Scale | 0 to 100 | 0 to 100 | Roughly 60 to 140 |
| Saturated color | Only R9 | Full circle | Full circle |
Reading the Color Vector Graphic
The graphic is a wheel made of 16 hue bins. Each bin carries an arrow. Arrow length shows the shift in chroma, arrow direction shows the hue rotation. A fan of outward arrows means the source over-saturates those hues. If the red bin points outward and the cyan bin points inward, the source will look warm and dense on wood and skin and flat on cool tones.
Three Numbers Worth Buying
1. Rf tells you how close the average is. Ask for a value and a tolerance, not a claim.
2. Rg tells you whether colors gain or lose punch. Rg near 100 with high Rf is the safe neutral.
3. The graphic, or at least the red and skin-tone bins, tells you what to expect on the products the light will actually show.
Where This Changes a Decision
Retail and museum work lives on hue accuracy, so Rf above 90 with Rg between 95 and 105 is a reasonable target, and the red bin matters more than the average. Industrial and warehouse work rarely needs any of it; uniformity and glare control dominate, and Rf 70 with a stable CCT is enough. Horticultural work is the opposite case again, where the whole point is to move energy outside the visible reference, so color metrics describe a side effect rather than a goal.
Two practical checks are worth running before a specification is frozen. First, confirm the metric version, because TM-30-15, TM-30-18 and TM-30-20 use updated sample sets and calculated values. Second, confirm the test condition, because a value measured at 25C with a specific driver does not transfer automatically to a fixture running at 85C case temperature.
FAQ
Does TM-30 replace CRI? Not on paper. Most compliance programs still accept CRI, and both can be reported together. TM-30 is the better engineering tool when hue matters.
Can a source have high Rf and low Rg? Yes. A narrow phosphor mix can average close to the reference while pulling saturation down. That combination reads as slightly gray next to a reference source.
Why does the reference change with CCT? In TM-30 the reference for fidelity is a daylight-like source at the same CCT as the test source, so the comparison is made at equal warmth rather than against one fixed white.
Related reading
