From the Studio
Two fixtures can both say "CRI 90" and make the same room look completely different. Here's what the number leaves out — and the three metrics we actually specify against.
By Ronny Esquenazi, Principal at Gridspec Lighting
Every cut sheet we review says some version of the same thing: "High CRI." "90+ CRI." "True color." It has become the lighting industry's equivalent of "organic" — technically meaningful, practically a marketing checkbox. And yet two fixtures with identical CRI ratings can make the same walnut millwork, the same terrazzo floor, or the same person's face look noticeably different under them.
That gap isn't an accident. It's built into how the number is calculated. If you're an architect, designer, or homeowner putting your name on a finished space, it's worth understanding what CRI actually measures — and what it quietly leaves out.
The general Color Rendering Index (Ra) is an average of how faithfully a light source renders eight test colors, labeled R1 through R8. Those eight samples are all moderately saturated, mid-tone colors. None of them is a strong, saturated red.
Saturated red has its own test sample — R9 — but it isn't part of the average. That matters, because red is everywhere in a well-designed interior: in skin tones, in warm woods, in leather, brick, terracotta, wine, food, and textiles. It's also one of the hardest parts of the spectrum for many LEDs to render well. A fixture can post a respectable CRI of 90 while its R9 lags far behind, and the result is a room where faces look slightly gray and rich materials look flat — even though the spec sheet says everything is fine.
This isn't just a designer's preference. California's Title 24 energy code, through its Joint Appendix 8 (JA8) qualification requirements for residential high-efficacy light sources, requires a CRI of 90 or higher and an R9 of 50 or higher. In other words, the state that writes some of the most influential lighting rules in the country decided that CRI on its own wasn't a sufficient guarantee of quality, and wrote the red-rendering requirement in separately.
Florida doesn't impose JA8, but the logic travels. If a code body felt the need to call out R9 explicitly, a spec that only lists CRI is leaving the most visible failure mode unaddressed.
CRI was developed decades ago, and the industry has since built a better tool. ANSI/IES TM-30 evaluates a light source against 99 color samples drawn from real-world object reflectances, rather than eight, and reports two primary numbers:
Rf (Fidelity) — how closely colors match a reference source, on a 0–100 scale. It's the conceptual successor to CRI, but far harder to game because it samples so much more of the spectrum.
Rg (Gamut) — whether colors are rendered more saturated (above 100) or more muted (below 100) than the reference. This is the number CRI simply can't tell you. A light source can have high fidelity and still lean slightly vivid or slightly dull, and that tilt is often exactly what makes a space feel warm and alive versus clinical.
TM-30 also groups its samples into 16 hue bins and plots them on a color vector graphic, so you can see where a source pushes or pulls color — whether reds are slightly oversaturated, greens slightly washed out, and so on. Hue bin 1 corresponds to red, which effectively gives you a more robust read on what R9 was trying to capture.
On our projects, CRI is the floor, not the specification. When we write a fixture schedule, we look for:
1. CRI and R9 together. We treat a published R9 value as non-negotiable, and we want to see it comfortably above the 50 that JA8 sets as a minimum for residential work — particularly in kitchens, dining rooms, primary baths, and anywhere people will see each other's faces.
2. TM-30 Rf and Rg. Where manufacturers publish full TM-30 reports, we read them. Fidelity tells us whether the light is honest; gamut tells us its personality. For hospitality and residential spaces, a slight lift in saturation is often the right call. For a gallery or a retail space selling color-critical product, we want fidelity first.
3. Consistency across the fixture package. Mixing sources with different rendering profiles in one room is one of the most common reasons a space feels "off" without anyone being able to say why. The cove, the downlights, and the decorative pendant should agree with each other.
South Florida interiors lean heavily on natural materials — white oak, limestone, travertine, terrazzo, warm plaster — and on spaces that transition from strong daylight to evening light. When the sun goes down, the artificial light has to carry those materials on its own. A source that renders reds and warm tones poorly will make an expensive palette look cheaper the moment the shades come down.
That's the difference between illuminating a room and engineering how it feels to be in it. The number on the box is where the conversation starts. The specification is where it gets decided.
Sources: California Energy Commission, Title 24 Part 6, Joint Appendix 8 (JA8) — Qualification Requirements for High Luminous Efficacy Light Sources; ANSI/IES TM-30, IES Method for Evaluating Light Source Color Rendition (Royer, "Tutorial: Background and Guidance for Using the ANSI/IES TM-30 Method," LEUKOS).
— Ronny Esquenazi, Principal, Gridspec Lighting
This article is general information, not engineering, legal or other professional advice, and may not reflect current codes, products or pricing. Consult your project’s licensed professionals before relying on it. See our Terms of Use.
