A requirement such as “90 CRI minimum” is a sensible starting point for an interior where people will judge finishes, food, merchandise, or skin tones. It is not, however, a complete description of what a light source will do to color. Two LED products can have the same CRI and make particular hue families look noticeably different.
ANSI/IES TM-30 gives a specifier more of the information needed to see that difference. It evaluates color rendition with 99 color evaluation samples and reports both average fidelity and average gamut, alongside graphics that show where color shifts occur. The U.S. Department of Energy’s TM-30 tutorial is a useful primary reference when a project team needs the full method and its intended use.
Start with what CRI does—and does not—say
The familiar general Color Rendering Index is commonly written as CRI or Ra. It compares a test source with a reference illuminant at the same correlated color temperature (CCT) and summarizes fidelity with one number. A higher value means the evaluated test colors match the reference more closely on average.
That makes CRI a practical baseline in a finish schedule or fixture legend. It is widely published, easy to compare, and still matters when a product has no TM-30 report. But it is an average based on a limited set of color samples. It cannot tell the reader whether a source tends to mute reds, push greens, or make one area of the hue circle more saturated while another is less saturated. It also cannot establish that every object will look “better”: color appearance depends on the source spectrum, the object, its surroundings, and the observer.
Use CRI to eliminate unsuitable options, then request more evidence when color is central to the visual task. That is especially worthwhile for downlights lighting art, food, or faces; linear lights washing a long material surface; and LED panel lights that form the ambient layer throughout a workplace.
The three TM-30 outputs to read first
TM-30 does not replace design judgement with a new single-number threshold. It gives a family of measures and graphics so the project team can match the source to its color-rendition intent. A product report normally makes these three outputs the quickest place to begin.
Rf: average fidelity
The TM-30 fidelity index, Rf, runs from 0 to 100. Like CRI, it is about average agreement with the reference illuminant; 100 represents no difference for the method’s color samples. Unlike CRI, TM-30 uses 99 Color Evaluation Samples. The larger, purpose-built sample set is one reason an Rf value is not interchangeable with a CRI value—even when both numbers appear high.
For a project whose goal is faithful appearance, Rf is the logical first screen. It is not a license to write “Rf equals CRI” in a specification or to translate one threshold mechanically into the other. Ask manufacturers for the report generated for the actual LED/driver/CCT combination being proposed.
Rg: average gamut or saturation change
The gamut index, Rg, describes the average area of the color gamut relative to the reference. An Rg of 100 means no average change in gamut area. Values above 100 indicate more average saturation; values below 100 indicate less average saturation. Rg is not a universal “higher is better” score: a grocery display, a hospitality interior, and a color-critical review space can reasonably have different design intents.
Read Rf and Rg together. A source may have decent average fidelity yet make colors more vivid or more muted overall. The DOE tutorial emphasizes that no one color-rendition measure identifies the best light source for every application; the intended visual outcome needs to be stated first.
Color Vector Graphic: where the average hides a problem
The Color Vector Graphic divides color space into 16 hue bins. Its shape and direction show the hue and chroma shifts that an average can conceal. This is the part of a TM-30 report to examine when a material palette is sensitive to a particular color family.
For example, if a retail concept depends on saturated red packaging, do not stop at a CRI or Rf headline. Review the vector graphic and, where relevant, the hue-bin and color-sample measures with a lighting designer and manufacturer. Request a physical mock-up with the actual finishes, because a graphic is a screening and communication tool—not a substitute for observing the project materials under the proposed light.
Write the design intent before the number
The most reliable color-rendition specification says what the space needs people to perceive. That keeps a project from treating a high score as an end in itself.
| Application question | Evidence to request |
|---|---|
| Must finishes, artwork, or skin tones look close to a reference condition? | A TM-30 report for the selected CCT, with Rf and the Color Vector Graphic; a mock-up with representative materials. |
| Should merchandise or food read as vivid, without an uncontrolled color cast? | Rf, Rg, the Color Vector Graphic, and a sample viewing under the proposed product. |
| Is the lighting primarily ambient office or circulation light? | The published CRI baseline, CCT, and a TM-30 report when color-sensitive finishes or brand standards make rendition important. |
This approach also keeps color rendition separate from adjacent, non-interchangeable requirements. CCT does not determine fidelity. Dimming performance, optical control, flicker behavior, illuminance, glare, and daylight response need their own review. For screen-intensive spaces, pair the color decision with a room-level glare assessment rather than assuming a color metric addresses UGR.
A practical submittal checklist
When color rendition matters, ask each bidder to provide:
- The specified CCT and the published CRI for the offered configuration.
- A TM-30 report for that exact light source configuration, including Rf, Rg, and the Color Vector Graphic.
- Any supporting color-sample or hue-bin data relevant to the materials in the project.
- Confirmation that replacement fixtures or alternate drivers retain the submitted color-rendition performance.
- A mock-up plan using the actual finish samples, viewed at the intended light level and geometry.
Avoid copying a generic “TM-30 compliant” claim into a schedule. TM-30 is a method for evaluating and communicating color rendition, not a pass/fail product certification. Write the project’s visual objective, identify the evidence needed to judge it, and leave room for the design team to review a real sample.
The takeaway
CRI remains useful as a simple, widely available floor. TM-30 becomes valuable when the project needs to understand how a spectrum changes color, not just whether it clears a broad average. Read Rf for average fidelity, Rg for average gamut change, and the Color Vector Graphic for hue-specific shifts. Then confirm the result with the finishes, objects, and viewing conditions that the project actually contains.
For deeper technical context, consult the DOE color-and-spectrum resources and the ANSI/IES TM-30 tutorial. They explain the method’s scope and why a deliberate color-rendition intent is more useful than a one-number rule.