Why Color Quality Matters

Walk into any well-designed space โ€” a luxury hotel lobby, a high-end boutique, a fine-dining restaurant โ€” and you'll feel the difference that good color quality makes. Skin tones look healthy, food looks appetizing, materials look rich, and the overall environment feels premium. Walk into a space with poor color quality, and everything feels flat, washed out, and cheap โ€” even if the finishes are expensive.

Color quality is one of the most important parameters in lighting specification, and yet it's also one of the most misunderstood. Most specifiers know about CRI (Color Rendering Index) and can tell you that CRI 90 is better than CRI 80. But there's so much more to color quality than just the CRI number.

Two LED fixtures with the same CRI 90 rating can produce dramatically different visual results. One might make reds look vibrant and skin tones look healthy, while the other makes reds look muddy and people look washed out. The difference is in the details โ€” the individual R values, the TM-30 metrics, the color binning, and the spectral power distribution.

This guide goes beyond the basic CRI number to explain the full picture of color quality in architectural lighting. You'll learn what CRI actually measures, where it falls short, what TM-30 adds, why R9 and R13 matter, how MacAdam ellipses affect consistency, and โ€” most importantly โ€” how to write color quality specifications that actually deliver the results you want.

CRI (Ra) Explained โ€” and Its Limitations

The Color Rendering Index (CRI) is the most widely used metric for describing how accurately a light source renders colors. It was developed in the 1960s and 70s by the CIE (International Commission on Illumination) and has been the industry standard ever since.

How CRI Is Calculated

CRI measures how accurately eight standard test color samples (R1 through R8) appear under a given light source compared to how they would appear under a reference source (daylight for cool white light, a blackbody radiator for warm white light).

Each R value is scored from 0 to 100, with 100 being perfect color rendering. The eight values are averaged to produce Ra โ€” the "general color rendering index" that's commonly referred to as just "CRI."

Ra
Average of R1-R8
The standard "CRI" number. Average color rendering across 8 pastel test colors.
R9
Saturated Red
Not included in Ra. Critical for reds, skin, food, and retail.
CRI 80
Commercial Baseline
Minimum for most commercial applications. Good but not great.

The Limitations of CRI

While CRI is useful as a general indicator, it has significant limitations that are especially problematic with LED technology:

  • Only 8 test colors, all pastel: CRI uses only eight test colors, all of which are low-to-medium saturation. It doesn't test saturated colors like deep red, green, or blue at all.
  • Developed for fluorescent, not LED: CRI was developed when fluorescent lighting was the dominant technology. LEDs have very different spectral distributions that can score well on CRI but still look poor to the human eye.
  • Average hides problems: Because CRI is an average of eight values, a light source can have a high CRI but still render certain colors very poorly. A CRI 90 light might have R9 (saturated red) of 0 or even negative.
  • No measure of saturation shift: CRI measures color accuracy (fidelity) but doesn't tell you whether colors appear more or less saturated than they should.
  • CIE 1964 color space: CRI uses an outdated color space that doesn't correlate well with human perception of color differences.

Common mistake: Specifying "CRI 90" and assuming that guarantees good color quality. Two different LED products with CRI 90 can look dramatically different. Always look beyond the CRI number and check the individual R values โ€” especially R9 โ€” and if possible, the TM-30 data.

TM-30-15: Rf and Rg, Why It's More Accurate

TM-30-15, published by the IES (Illuminating Engineering Society) in 2015, is the most advanced and comprehensive color rendering metric currently available. It was developed specifically to address the limitations of CRI and to provide a more complete picture of color quality.

What Makes TM-30 Better

TM-30 improves on CRI in several important ways:

  • 99 color evaluation samples (CES): Instead of 8 pastel colors, TM-30 uses 99 color samples spanning the full range of hues, saturations, and values. This gives a much more complete picture of color rendering.
  • Updated color space: TM-30 uses the CAM02-UCS color space, which correlates much better with human color perception than the CRI color space.
  • Two values instead of one: TM-30 provides both Rf (fidelity) and Rg (gamut), giving you more information than just a single number.
  • Vector graphic representation: TM-30 includes a graphical representation of how each hue is shifted in both hue and saturation, making it easy to see exactly how colors are affected.

Understanding Rf (Color Fidelity)

Rf is the TM-30 equivalent of CRI Ra โ€” it measures how accurately colors are rendered compared to the reference source. Like CRI, it's scored from 0 to 100, with 100 being perfect.

Because Rf uses 99 test colors instead of 8, it's a much more robust and reliable measure of overall color rendering. A light source with Rf 90 will consistently render colors well across the full spectrum, not just on the 8 CRI test colors.

Understanding Rg (Color Gamut)

Rg is the second TM-30 value, and it measures whether colors appear more or less saturated under the test light source compared to the reference.

  • Rg = 100: Colors appear with the same saturation as the reference โ€” perfectly natural.
  • Rg > 100: Colors appear more saturated (vivid, enhanced). Rg 110 means colors are about 10% more saturated.
  • Rg < 100: Colors appear less saturated (muted, washed out). Rg 90 means colors are about 10% less saturated.

Rg is important because two lights with the same Rf (fidelity) can have very different subjective appearances depending on their Rg. A light with Rf 90 and Rg 100 renders colors accurately and naturally. A light with Rf 90 and Rg 110 also renders colors accurately, but they appear more vivid and saturated โ€” which might be desirable for retail or restaurants but not for galleries or healthcare.

TM-30 target ranges by application: For general architectural lighting, aim for Rf โ‰ฅ 90 and Rg between 95 and 105. For retail and restaurants, slightly elevated gamut (Rg 100-110) can make products and food look more appealing. For galleries and museums, aim for Rf โ‰ฅ 95 and Rg as close to 100 as possible for the most accurate color rendering.

Special R Values: R9, R13, and Why They Matter

In addition to the eight R values that make up CRI (R1-R8), there are seven supplementary R values (R9-R15) that measure specific saturated and special colors. These aren't included in the average CRI calculation, but they're often more important than the average for real-world applications.

R9

Saturated Red โ€” The Most Important Special R Value

R9 measures how accurately a saturated red color is rendered. It's critically important because red is everywhere in the built environment: red bricks, wood tones, skin tones, lips, food, fabric, artwork, logos. A light source with poor R9 makes all of these look dull, brown, and lifeless.

Many budget "CRI 90" LEDs have surprisingly low R9 values โ€” sometimes zero or even negative. This is because it's relatively easy to make an LED with good average CRI by tuning the phosphor, but getting good saturated red rendering requires either a red LED chip (RGB/RGBA) or a special high-CRI phosphor formulation, both of which add cost.

R9 ValueQuality LevelSuitable Applications
< 0PoorNot recommended for any architectural application
0-30FairIndustrial, warehouse, utility lighting only
30-50GoodStandard commercial, basic retail, office
50-70Very GoodPremium retail, hospitality, residential
70-90ExcellentLuxury retail, high-end hospitality, healthcare
90+Museum / Gallery GradeGalleries, museums, high-end fashion retail
R13

Skin Tone โ€” Critical for People Spaces

R13 measures skin tone rendering. It's important for any space where people are present โ€” hospitality, retail, offices, healthcare, residential. Good R13 makes people look healthy and vibrant; poor R13 makes people look sallow, gray, or washed out.

R15

Japanese Skin Tone โ€” Often Overlooked

R15 measures a second skin tone sample (Japanese complexion). It's especially relevant for projects in Asia or with diverse populations. Good R15 indicates that the light source renders a range of skin tones well, not just the one tested in R13.

Other Notable R Values

  • R10 (saturated yellow): Important for warm wood tones, yellow fabrics, and gold/brass finishes.
  • R11 (saturated green): Important for landscape lighting, plants, green materials, and healthcare.
  • R12 (saturated blue): Important for blue fabrics, glass, water features, and cool-toned materials.

MacAdam Ellipses and Color Consistency

Color quality isn't just about how accurately a single fixture renders colors โ€” it's also about how consistently multiple fixtures match each other. Nothing looks more unprofessional than a row of downlights where some look slightly pink, some slightly green, and some slightly yellow, even though they're all supposedly the same "3000K."

How Color Binning Works

LEDs are manufactured in batches, and even within the same batch, there's natural variation in color temperature and tint. To manage this, manufacturers sort (bin) LEDs into groups based on their measured color coordinates on the CIE 1931 chromaticity diagram. Tighter bins mean more consistent color โ€” and higher cost.

The size of these bins is measured in MacAdam steps (also called SDCM โ€” Standard Deviation of Color Matching), based on research by David MacAdam in the 1940s. MacAdam found that the human eye can just barely detect a color difference of 1 step.

MacAdam Steps (SDCM)Visual PerceptionQuality Tier
1-stepColor difference is imperceptibleMuseum / gallery grade (rare, very expensive)
2-stepBarely perceptible side-by-sidePremium / luxury standard
3-stepVisible on close inspectionStandard commercial quality
5-stepObvious color differenceBudget / economy grade
7-step+Very obvious, unacceptableShould be avoided for architectural lighting

Specification tip: For premium architectural projects, specify 2-step MacAdam maximum. For standard commercial projects, 3-step is acceptable. Never specify 5-step or looser binning for visible lighting applications โ€” the color variation will be immediately obvious and will make even an expensive installation look cheap. Also, ask the manufacturer about bin mixing โ€” some manufacturers mix bins within a single production run, which can cause visible variation even within a 3-step spec.

Color Quality in Different Applications

The right level of color quality depends on the application. Higher is always better, but there's a cost trade-off, and the marginal benefit of CRI 98 vs CRI 90 may not be worth the cost in every situation.

Retail and Luxury Retail

In retail, color quality directly impacts sales. Customers need to see products in their true colors to make purchasing decisions, and good color quality makes merchandise look more appealing and valuable.

  • CRI: 90 minimum, 95+ for luxury
  • R9: โ‰ฅ 50 minimum, โ‰ฅ 70 for fashion and cosmetics
  • TM-30 Rf: โ‰ฅ 90
  • TM-30 Rg: 100-110 (slightly elevated gamut enhances merchandise)
  • Color consistency: 3-step MacAdam minimum, 2-step for premium
  • CCT: 2700K-3500K depending on brand identity

Healthcare and Hospitals

In healthcare, color quality is not just an aesthetic issue โ€” it can affect clinical outcomes. Doctors and nurses need to accurately assess skin tone, tissue color, wound appearance, and medication colors.

  • CRI: 90 minimum for patient areas
  • R9: โ‰ฅ 50 minimum, โ‰ฅ 70 preferred (critical for skin and blood assessment)
  • R13: High priority for accurate skin tone assessment
  • TM-30 Rf: โ‰ฅ 90
  • TM-30 Rg: 95-105 (accurate โ€” no artificial enhancement)
  • Color consistency: 3-step MacAdam minimum
  • CCT: 3500K-4000K for clinical areas, 2700K-3000K for patient rooms

Hospitality and Hotels

In hospitality, color quality contributes directly to the guest experience. Warm, flattering light makes guests feel comfortable and makes the space feel luxurious. Skin tone rendering (R13) and red rendering (R9) are especially important.

  • CRI: 90 minimum, 90+ standard for luxury hotels
  • R9: โ‰ฅ 50 minimum, โ‰ฅ 70 for luxury
  • R13: Excellent skin tone rendering is essential
  • TM-30 Rf: โ‰ฅ 90
  • TM-30 Rg: 98-105 (natural to slightly enhanced)
  • Color consistency: 3-step minimum, 2-step for premium spaces
  • CCT: 2700K standard for hospitality, 3000K for contemporary brands

Galleries and Museums

Galleries and museums have the most demanding color quality requirements. Artwork must be displayed with accurate colors so visitors see it as the artist intended.

  • CRI: 95+ minimum, 98+ for museum-grade
  • R9: โ‰ฅ 90 minimum
  • All R values: All 15 R values should be high
  • TM-30 Rf: โ‰ฅ 95
  • TM-30 Rg: 98-102 (as close to 100 as possible)
  • Color consistency: 2-step MacAdam minimum
  • CCT: 2700K-3500K depending on the collection type

Offices

In offices, color quality affects employee comfort, productivity, and wellbeing. While offices don't need gallery-level color rendering, poor color quality contributes to a cheap, unpleasant environment.

  • CRI: 80 minimum for standard offices, 90 for premium
  • R9: โ‰ฅ 0 minimum (avoid negative R9!), โ‰ฅ 30 for premium
  • TM-30 Rf: โ‰ฅ 80 standard, โ‰ฅ 90 premium
  • TM-30 Rg: 95-105
  • Color consistency: 3-step MacAdam minimum
  • CCT: 3500K-4000K

How to Specify Color Quality Correctly

A good color quality specification is specific, measurable, and verifiable. "High CRI" is not a specification โ€” it's a marketing claim. Here's how to write color quality specifications that actually deliver what you want.

The Complete Color Quality Specification

Sample Color Quality Specification

For a premium retail project

1. Color Rendering Index (CRI): Minimum CRI (Ra) 90. All individual R values (R1-R15) shall be provided. R9 shall be โ‰ฅ 50, R13 shall be โ‰ฅ 70.

2. TM-30: TM-30-15 values shall be provided. Rf โ‰ฅ 90, Rg = 100-110. TM-30 vector graphic shall be provided upon request.

3. Color Consistency: LED color binning shall be 3-step MacAdam (SDCM) maximum, 2-step typical. All fixtures within a single order shall be supplied from the same or adjacent bins. Manufacturer shall provide binning data upon request.

4. Color Temperature: 3000K ยฑ 100K (at nominal current and 25ยฐC ambient). Measured in accordance with LM-79.

5. Testing: All color parameters shall be measured per IES LM-79 at the rated operating current and ambient temperature. Third-party test report shall be provided upon request.

Key Specification Principles

  • Specify minimums, not averages: "Minimum CRI 90" is clear; "CRI 90" is ambiguous.
  • Include R9 and R13: Don't rely on just the average CRI. Specify minimum values for critical R values.
  • Specify MacAdam steps: Always specify the maximum color consistency step. 3-step is standard commercial; 2-step is premium.
  • Ask for test data: Require LM-79 test reports that include all color parameters.
  • Consider TM-30: For premium projects, add TM-30 Rf and Rg requirements to the specification.
  • Request samples: Always evaluate actual samples before final approval. Numbers on a datasheet can't replace seeing the light in person.

Common Myths and Misconceptions

Myth: "CRI 90 means the light has good color quality"

Not necessarily. CRI 90 tells you the average of 8 pastel colors is good, but it doesn't tell you anything about saturated colors (especially reds), skin tones, or whether colors appear over- or under-saturated. Always check R9 and R13 along with the CRI number, and ideally the TM-30 data.

Myth: "Higher CRI means lower efficacy"

This was true in the early days of LED, but it's much less true today. Modern high-CRI LEDs are nearly as efficient as standard CRI 80 LEDs. The difference is typically 5-10%, not the 30-40% it used to be. For most projects, the quality benefit of CRI 90 far outweighs the minor efficiency cost.

Myth: "All CRI 90 LEDs are basically the same"

Far from it. Two different CRI 90 LEDs can have very different spectral power distributions and very different visual appearances. Some use a "boosted red" approach that makes reds pop but can make other colors look unnatural. Some have a more balanced spectrum that's more natural overall. The only way to know is to see them side by side.

Myth: "Warm white light automatically makes spaces feel luxurious"

Warm light helps, but it's not the whole story. A warm light with poor color quality (low CRI, bad R9) can make a space feel dingy and cheap. A cooler light with excellent color quality can feel clean, modern, and sophisticated. The color quality matters at least as much as the color temperature.

Final tip: When evaluating color quality, there's no substitute for seeing it in person. Request samples of the actual fixtures you're considering and view them side by side in the actual space, with actual materials (wood finishes, fabrics, paint colors). Numbers on a datasheet are useful for initial screening, but the final decision should always be based on a visual evaluation.

Color quality is one of the most impactful yet most under-specified aspects of architectural lighting. A fixture with great color quality can make an ordinary space feel premium, while poor color quality can make an expensive space feel cheap. By going beyond the basic CRI number and specifying R values, TM-30 metrics, and MacAdam step binning, you ensure that the lighting in your project delivers the color quality you and your clients expect.

If you're working on a project with demanding color quality requirements โ€” retail, hospitality, healthcare, or gallery โ€” explore our high-CRI lighting collections or contact our technical team for color quality specification support and sample evaluation.