CRI vs TLCI: Understanding Color Accuracy in LED Lights

If you've shopped for LED video lights, you've seen specs like "CRI 95+" or "TLCI 98" stamped on the box. Both numbers are attempts to describe how faithfully a light reproduces colors compared to a reference source — but they were developed for different purposes, by different standards bodies, and they answer different questions. This explainer covers what each metric measures, where each one is and isn't useful, and the most common misconceptions that surface in product marketing.

What CRI Actually Is

The Color Rendering Index (CRI) is a metric defined by the International Commission on Illumination (CIE) — see CIE Publication 13.3, "Method of Measuring and Specifying Colour Rendering Properties of Light Sources." CRI compares a test light to a reference illuminant of the same correlated color temperature (a Planckian black-body radiator below 5000 K, or a defined daylight illuminant above) and scores how accurately the test light renders a set of standardized color samples.

The CIE 13.3 method defines 14 reference samples, R1 through R14 (R15, an Asian skin-tone sample, was added later by some testing protocols and isn't part of the original 1995 CIE 13.3). Each sample produces an individual rendering score (R1, R2, etc.). The headline number on a product datasheet — usually written just as "CRI 95" or "Ra 95" — is Ra, the general color rendering index, which is the arithmetic mean of only the first eight samples (R1–R8). R1–R8 are pastel, low-to-moderate-saturation colors.

The samples that aren't included in the Ra average — R9 through R14, plus R15 where added — are commonly called the "extended" or "special" CRI values. They cover saturated reds (R9), saturated yellows, greens, blues, plus skin-tone references and, in extended protocols, foliage. R9 (saturated red) is the one to watch for video and photography because human skin tone, blood-flush, and many fabrics depend on accurate red rendering, and a light can have a high Ra (R1–R8 average) while scoring poorly on R9. Reference: CIE Publication 13.3-1995 (CIE 13.3).

What TLCI Is, and How It Differs

The Television Lighting Consistency Index (TLCI) is a separate standard developed by the European Broadcasting Union and published as EBU Tech 3355. Unlike CRI, which models how the human eye perceives color rendering, TLCI models how a video camera sees a light source. The TLCI test method uses a 24-patch ColorChecker-style chart, illuminates it with the test source, simulates capture by a reference camera (with a defined spectral response), applies a defined color-correction stage, and then compares the result to a reference image rendered under the reference illuminant. The output is the TLCI number on a 0–100 scale (EBU Tech 3355).

The practical implication: TLCI accounts for the camera's spectral sensitivity and the typical broadcast color-correction workflow, so it can flag problems CRI misses (and occasionally pass lights CRI marks down). Both metrics are useful; neither is a complete substitute for the other, and high-end lighting datasheets increasingly publish both.

How to Read the Numbers

Two lights with the same CRI Ra can have very different R9, and two lights with similar Ra can have meaningfully different TLCI. When evaluating a light for video work:

• CRI Ra — average of R1–R8. A useful summary, but not the whole picture.
• R9 — saturated red. Critical for skin tones. A high-Ra light with R9 in the 30s or 40s will render skin poorly.
• R12, R13, R14, R15 — saturated blue, light skin, and (where reported) Asian skin-tone references.
• TLCI — when reported, useful as a complementary check oriented toward video capture.
• SSI (Spectral Similarity Index) and TM-30 Rf/Rg — newer metrics that more granularly describe how a source's spectral output compares to a reference. Increasingly common on professional datasheets.

A rough working guideline used widely in the production community: for general content creation, look for CRI Ra ≥ 90 and R9 ≥ 80; for broadcast and color-critical narrative work, CRI Ra ≥ 95 with R9 ≥ 90 is a more comfortable threshold (Sound on Sound and B&H Explora video resources publish similar guidance regularly).

Common Misconceptions

"CRI only tests 8 colors." Not quite. The CIE method defines 14 reference samples; what you usually see on a box (Ra) is the average of only the first eight. The full set is calculated and reported separately when manufacturers choose to publish it.

"A high CRI guarantees accurate skin tone." Not by itself. Because Ra averages only R1–R8, a light can post a high Ra and still render saturated reds poorly, which directly affects skin tones. Look for R9 specifically.

"TLCI replaces CRI." They were developed by different standards bodies (CIE vs. EBU) and answer different questions (perceptual rendering vs. camera-captured rendering). Many professional datasheets now report both.

"Daylight balance and CRI are the same thing." They aren't. Color temperature (e.g., 5600 K vs. 3200 K) describes where on the warm-to-cool axis a light sits. CRI describes how faithfully the light renders the reference samples at that temperature. A 5600 K source can have great or terrible CRI; the two ratings are independent.

"You can fix bad color in post." Color correction can shift hue, saturation, and white balance, but it cannot recover wavelength information that the source never emitted. If a light's spectrum has gaps — typical of inexpensive single-phosphor LEDs — those gaps show up as missing color information that no grading tool can manufacture.

"More expensive always means better color." Generally true at the extremes, but not a reliable rule in the middle. Some premium lights spend their cost on wireless control, build, or output rather than spectral quality. Pricing for high-CRI panels at retailers like B&H spans a wide range — entry-level panels with credible CRI 95+ ratings are widely available, while top-tier broadcast fixtures cost considerably more (B&H LED light catalog).

Why "Cheap" Spectra Often Look Off

A common observation in the production community is that some inexpensive LED panels render skin and foliage with a perceptible green or magenta cast on camera. The technical reason is that low-cost white LEDs typically use a blue pump LED with a single-phosphor down-conversion, which leaves visible gaps in the spectral power distribution — particularly in the red region. That spectral gap shows up as a low R9, and on a camera sensor it can manifest as a hue cast that's hard to fully neutralize with white balance alone. This is a generic property of single-phosphor LED architectures, not a claim about any specific brand or product. Reviews from outlets like Sound on Sound (for their lighting coverage in the post/video adjacent space) and B&H Explora's lighting articles regularly discuss spectral power distribution charts for specific fixtures.

When This Affects You

If you're shooting documentary, narrative, broadcast, corporate, or product work, color rendering is a baseline requirement. If you're streaming or recording talking-head video where the subject's face is the main element of the frame, R9 in particular will affect how natural your skin tone reads on camera. If you're using LED practicals on a film set, TLCI ratings (or SSI / TM-30 data when available) are the cleaner check because they model camera capture rather than human perception.

For most non-color-critical applications — task lighting, ambient room light, or B-roll where color isn't tightly graded — Ra in the high 80s with no obvious tint is generally fine. The bar rises sharply once skin tones are on camera.

Sources & Citations

1. International Commission on Illumination — CIE Publication 13.3-1995, "Method of Measuring and Specifying Colour Rendering Properties of Light Sources." cie.co.at
2. European Broadcasting Union — EBU Tech 3355, "Method for the Assessment of the Colorimetric Properties of Luminaires (TLCI / TLMF)." tech.ebu.ch
3. Illuminating Engineering Society — TM-30 method overview. ies.org
4. Sound on Sound — production audio and adjacent video/lighting coverage. soundonsound.com
5. B&H Photo Video Explora — LED lighting buying guides and spectral discussion. bhphotovideo.com/explora/video