CRI LED, TLCI and R9: color quality

LED CRI and color quality: why color matters

Before delving into the specific metrics, it's essential to understand what we're talking about. Color quality, or color rendition, describes the effect of a light source on the apparent color of objects. It is not an intrinsic property of the light itself, but the result of the interaction between the emission spectrum of the source and the spectral reflectance properties of the objects.

A lamp – often compared to natural midday sunlight or a black body at the same color temperature – possesses a continuous and complete spectrum containing energy at all visible wavelengths. This allows it to illuminate any color, reflected by the object in its entirety, returning a faithful and rich perception.

Artificial sources, especially those with discrete spectra like LEDs, can have "gaps" or pronounced peaks in some spectral bands, leading to more or less marked distortions. Objectively and standardly evaluating these distortions is the goal of the indices we will analyze.

CRI (Color Rendering Index): definition, history

CRI, or Color Rendering Index, is the most historically used and internationally recognized metric. Developed by the Commission Internationale de l'Éclairage (CIE) and formalized in publication CIE 13.3 (1995), CRI provides a comparative assessment of a light source's ability to reproduce colors compared to a reference source.

It is essential to emphasize that LED CRI does not measure the "beauty" or "saturation" of colors, but their fidelity relative to a reference. A high CRI indicates that colors will appear very similar to how they would appear under the reference source, given the same chromatic adaptation conditions for the observer.

What is CRI? Technical definition and conceptual basis

CRI, an acronym for Color Rendering Index, is a quantitative index that measures the degree of correspondence between the perceived color of an object illuminated by a test source and the perceived color of the same object illuminated by a reference source, when both sources have the same correlated color temperature (CCT). The index is based on the concept of chromatic shift.

In practical terms, a series of standardized color samples (originally 8, extended to 14) are selected, and it is calculated how each of these samples shifts in the CIE 1964 (U*, V*, W*) color space when illuminated by the test source compared to the reference source. The greater the shift, the lower the rendition for that particular sample. The arithmetic mean of the values obtained for the first 8 samples (R1 to R8) provides the general CRI, called Ra (where "a" stands for "average").

The question "What is CRI?" is thus answered by a precise algorithm:

1) The CCT of the test source is determined;

2) A Planckian black body (for CCT < 5000K) or a CIE daylight spectrum (for CCT ≥ 5000K) at the same CCT is chosen as reference;

3) The chromatic coordinates of the samples under the test source and under the reference are calculated;

4) A transformation is applied to correct for chromatic adaptation differences (using the von Kries transformation);

5) The color difference ΔEi for each sample i is calculated. 6) Each ΔEi is converted into a special index Ri via the formula: Ri = 100 - 4.6 * ΔEi;

7) The Ra (or general CRI) is the arithmetic mean of R1, R2, ..., R8.

The CRI color samples: the special indices from R1 to R14

The choice of color samples is critical. The first 8 samples (R1-R8) are pastel colors of medium saturation, representative of common pigments. They are useful for a general assessment but can mask specific spectral deficiencies. For this reason, the CIE introduced 6 supplementary samples (R9-R14), more saturated, which test specific spectral regions.

The R9 index deserves special mention. Since many white LEDs (especially those based on blue+yellow phosphors) emit little in the deep red spectrum (around 630-660 nm), the R9 value is often very low (even negative) despite a high Ra. An LED with Ra 90 and R9 < 20 will render reds dull, greyish, and lacking vibrancy. For critical applications, specifying a minimum R9 value (e.g., R9 > 50 or R9 > 80) is an indispensable professional practice.

How is CRI calculated? The step-by-step process

Manual calculation of CRI is complex and requires spectroradiometric instrumentation and specialized software. However, understanding the logical flow is fundamental to correctly interpreting the result. The process unfolds in a chain of mathematical and chromatic operations:

1. spectral measurement: the spectral power distribution (SPD) of the test source in the 380-780 nm range is acquired.
2. CCT calculation: from the spectrum, the chromatic coordinates (x,y) are calculated and the Correlated Color Temperature (CCT) of the test source is determined.
3. reference choice: based on the CCT, the spectrum of the reference source (black body or CIE daylight) is mathematically generated.
4. sample coordinate calculation: for each of the 14 samples, whose spectral reflectance is known, the chromaticity and luminance coordinates under the test source and under the reference are calculated, in the CIE 1964 U*V*W* color space (a uniform space where distances approximately correspond to perceptual differences).
5. chromatic adaptation correction: since the human eye adapts to different color temperatures, a transformation (CIE CAT) is applied to the coordinates of the test source to simulate its vision in the same adaptation state as the reference.
6. color difference calculation (ΔE): for each sample i, the color difference ΔEi in the U*V*W* space between its appearance under the reference and under the (corrected) test source is calculated.
7. conversion to special indices (Ri): each ΔEi is converted into a special index Ri: Ri = 100 - 4.6 * ΔEi. The coefficient 4.6 scales the result so that a high-pressure sodium lamp, with poor color rendering, has a CRI around 25. A ΔE equal to 0 (no difference) gives Ri=100.
8. calculation of Ra (General CRI): the Ra value (or CRI) is the arithmetic mean of the first 8 special indices: Ra = (R1 + R2 + ... + R8) / 8.

The formula for the color rendering index, at its core, is therefore Ri = f(ΔEi) = 100 - k * ΔEi, where k is a normalization constant. The complexity lies entirely in the accurate calculation of ΔEi, which must account for all the psychophysical factors of color vision.

The limits of CRI and the birth of new metrics: TLCI and TM-30-18

Despite its widespread use, CRI has critical limitations, which emerged especially with the advent of LED sources. These limits have driven research towards more robust alternative metrics.

Criticalities of the CRI method: why a high CRI is sometimes not enough

The main criticisms of CRI are of a technical and perceptual nature:

• sample choice: the pastel samples (R1-R8) are not representative of real saturated colors. A source can have a high Ra but a very poor rendition of reds (R9) or greens (R11);
• reference source: the reference is always a black body or daylight spectrum, even for sources with very different spectra (e.g., multi-peak LEDs). This can lead to unfair evaluations;
• obsolete color space: the CIE 1964 U*V*W* color space has been superseded by more uniform spaces like CIELAB or CIELUV. The non-uniformity can weigh differently on different chromatic regions;
• lack of preference indicators: CRI measures fidelity, not preference. Studies show that observers often prefer a slight enhancement of saturation, especially in retail. A pure fidelity metric does not capture this aspect;
• problems with very high or low CCT sources: the method becomes unstable for CCTs far outside the 2500K-6500K range.

These criticalities have made it clear that Ra alone is an insufficient indicator for a complete professional assessment. It is necessary to examine the supplementary indices, primarily R9, and consider more modern metrics.

TLCI (Television Lighting Consistency Index): the standard for broadcast

With the transition of television from analog to digital and from SD to HD and 4K, the need for stringent chromatic control for studio lights became pressing. CRI, designed for human observation, did not account for the response of digital cameras. The European Broadcasting Union (EBU) therefore developed the TLCI (Television Lighting Consistency Index), standardized as EBU Tech 3353 and later also adopted by the CIE.

TLCI answers a specific question: "How will colors appear when captured by a standard camera and reproduced on a reference monitor?" It replaces the human observer with a standardized electronic camera model, simulating the entire chain of acquisition, signal processing, and display.

TLCI: what it is and how it works

The TLCI-2012 methodology (and the subsequent TLCI-2015) follows this scheme:

1. camera model: a mathematical model of an HD camera with spectral response characteristics of RGB filters defined by ITU-R BT.709 recommendation is used.
2. sample set: a set of 18 color samples (including skin tones, EBU logo colors, saturated colors), more representative of a television scenario, is used.
3. signal chain simulation: for each sample, the following is simulated:
   • the RGB response of the camera under the test source and under a D65 reference source;
   • the camera's automatic white balance (AWB) correction;
   • gamma correction and video signal encoding;
   • decoding and display on a calibrated reference monitor (also compliant with BT.709).
4. color difference calculation: the color difference ΔE (in CIELAB space) between the image of the sample under the test source and under the D65 reference, after the entire processing chain, is calculated;
5. TLCI value assignment: the color differences ΔEi are converted into a qualitative index (Qa) and then into a TLCI score on a scale from 0 to 100, approximated to 5 points. The conversion is such that:
   • TLCI ≥ 85: excellent. No color correction needed in post-production;
   • TLCI between 70 and 85: good. Minor corrections may be needed;
   • TLCI between 50 and 70: acceptable. Substantial corrections will be necessary;
   • TLCI < 50: insufficient. Even with corrections, results will be poor.

The fundamental difference between CRI and TLCI lies in the "detector": the chromatically adapted human eye for CRI, the standardized camera-monitor system for TLCI. For a lighting designer working in television, cinema, or video production, TLCI is a more reliable and direct parameter than CRI for predicting light behavior in front of the lens.

The TM-30-18 metric: the modern evolution of chromatic evaluation

To systematically address all criticisms of CRI, the North American Illuminating Engineering Society (IES) developed the TM-30-18 (IES Method for Evaluating Light Source Color Rendition). This standard is not presented as a single number, but as a set of values and graphs providing a multidimensional analysis. TM-30-18 introduces two main indices and visual tools:

• Rf (Fidelity Index): Fidelity Index. Similar to CRI, but based on 99 real color samples (fabrics, paints, natural materials, skin, foliage), a more modern color space (CAM02-UCS), and a reference source that is the average of many high-rendering real sources, not just black body/daylight. Rf ranges from 0 to 100.
• Rg (Gamut Index): Gamut Index. Measures the average change in saturation. An Rg value = 100 indicates that the test source, on average, does not alter saturation compared to the reference. Rg > 100 indicates an average increase in saturation, Rg < 100 a decrease. This distinguishes fidelity (Rf) from "vividness" (Rg).
• Graphical tools: provides a vector diagram showing, for 16 color tones (hue bins), whether the source tends to saturate or desaturate and shift the hue. Also provides a graph of the emission spectrum.

TM-30-18 represents the state of the art in color rendition evaluation, offering a much richer and more reliable informational framework than Ra alone. Although not yet widely reported in LED datasheets, its adoption is growing in the professional sector.

Luminous efficacy and energy efficiency: balancing with color quality

An indispensable chapter in choosing a professional LED is the relationship between color quality and efficiency. Often there is a trade-off between high LED CRI/TLCI and luminous efficacy (lumens per watt). Understanding this relationship is fundamental to optimizing the lighting project both in terms of visual performance and energy consumption.

Definition of luminous efficacy and energy efficiency

It's important to distinguish two often confused concepts:

• Luminous efficacy of a source: it is the ratio between the total luminous flux emitted (in lumens, lm) and the absorbed electrical power (in watts, W). Measured in lm/W. Formula: η = Φ / P, where η is luminous efficacy, Φ is luminous flux, P is absorbed power. Indicates how efficient a source is at converting electrical energy into visible light.
• Energy efficiency: a broader concept considering the entire system (source + driver + control system) and final use. It can refer to the total consumption of a building or an application. An efficient lamp (high lm/W) contributes to overall energy efficiency.
• Luminous flux (Φ): measure of the luminous power perceived by the human eye, weighted according to the standard photopic sensitivity curve V(λ). Measured in lumens (lm). Answers the question: "How much light does this source emit overall?"

The relationship between LED spectrum, CRI, and luminous efficacy

To understand the trade-off, one must look at the physics of the white LED. Most white LEDs use a chip that emits blue light (around 450 nm) that excites a yellow phosphor (like YAG:Ce) placed on top. The mix of residual blue and phosphor yellow light gives the sensation of white. This system is very efficient because it converts energy well. However, the resulting spectrum essentially consists of two peaks (blue and a broad yellow-green band), with little deep red. This produces a low R9 and a CRI generally in the order of 70-80, but very high luminous efficacy (up to 200 lm/W for the best laboratory examples).

To increase LED CRI and particularly R9, manufacturers must modify the phosphor composition. By adding red phosphors (e.g., nitrides or oxynitrides doped with Eu2+) or using mixes of green and red phosphors ("multi-phosphor" or "violet/blue pump + multiple phosphors" approach), the spectrum in the red region is filled and the rendition of all samples improves. However, these additional phosphors often have a lower conversion efficiency than the classic yellow YAG:Ce phosphor, and they absorb part of the light emitted by other phosphors (re-conversion). The result is an overall loss of luminous efficacy.

Practical comparison: luminous efficacy as a function of CRI and R9

The following table illustrates, indicatively, the typical trend of luminous efficacy for LED COB (Chip-on-Board) modules of similar power (around 3000K CCT) varying with color rendering indices. Values are representative of the professional market.

The choice therefore becomes a balance between project needs: Is it more important to maximize energy efficiency (and reduce operating costs and number of light points) or to maximize color quality (and thus the visual experience and enhancement of environments and objects)? In professional projects, it is standard practice to calculate the total required flux and, based on the chosen lm/W, size the installed power. A lower lm/W may mean more luminaires or more powerful ones, with a possible increase in initial costs and contracted power.

Practical applications and guidelines for selection

We now provide an operational synthesis to guide the selection of LED sources based on the application context, integrating all discussed parameters.

Guidelines for specific sectors

1. Retail and visual merchandising

Objective: enhance products, make colors attractive and faithful, create pleasant atmospheres.

• Clothing and fabrics: a CRI > 90 and R9 > 50 are fundamental. High R9 ensures vibrant reds (e.g., t-shirts, red clothes, warm skin tones on mannequins). Also consider R13 (skin tones). TM-30-18 with Rg slightly >100 can increase appeal.
• Jewelry and watchmaking: CRI > 95, R9 > 90. Rendering precious stones (emeralds, rubies, sapphires) and metals (gold, platinum) requires a full and uniform spectrum. Beware of spectral peaks that can create unnatural reflections.
• Food and supermarkets: critical for meat, fruit, vegetables. CRI > 90, R9 > 80 is essential to make meat appear fresh and red, tomatoes ripe, vegetables lively. Warm CCTs (2700K-3000K) for deli sections, cooler (4000K) for produce sections.

2. Museums, art galleries, and cultural heritage

Objective: maximum color fidelity to respect the artist's intent, minimize photochemical damage.

• Paintings and artworks: CRI > 95, R9 > 90. Prefer LEDs with continuous or nearly continuous spectra, avoiding pronounced peaks. TM-30 Rf is an excellent indicator. Consider sources with CRI 98-99 (like daylight spectrum reproductions) for the most critical applications.
• Accent lighting: beyond color, control light distribution (beam angle) and absence of UV/IR for conservation.

3. Broadcast, cinema, and professional photography

Objective: accurate and consistent chromatic reproduction through cameras and film.

• TV studio and video production: the primary parameter is TLCI. Look for lights with TLCI ≥ 85 (labeled as "Class A"). Also check CCT consistency between different units (to avoid color casts between lights).
• Cinema: in addition to high TLCI, precise CCT (e.g., 3200K for tungsten, 5600K for daylight) and a high fidelity index (CRI/TM-30 Rf) are often required. The ability to dim without color shift (stable CCT with intensity variation) is crucial.

4. Offices, schools, and healthcare

Objective: visual comfort, productivity, accuracy in tasks.

• Offices and classrooms: CRI > 80 is considered the minimum. For long working hours, CRI > 85-90 reduces visual fatigue and improves detail perception (e.g., charts, color codes). Combine with a suitable CCT (4000K for concentration).
• Healthcare (diagnosis and examination rooms): in medical settings, the rendition of skin and its tones is vital. CRI > 90, with particular attention to R9 (for erythema, cyanosis) and R13/R15 (skin tones). Standards like IEC 60601-2-41 for surgical lighting prescribe very stringent CRI.

LED CRI and chromatic evaluation: for truly professional applications

The evaluation of color quality of LED sources for professional applications can no longer rely solely on the CRI (Ra) value. An advanced professional specification must include a series of parameters such as:

1. Correlated Color Temperature (CCT): the base white tone (e.g., 2700K, 3000K, 4000K).
2. General Color Rendering Index (CRI - Ra): a first indicator of fidelity, but not to be considered in isolation.
3. Special CRI Indices, particularly R9 (Saturated Red): fundamental to understand performance in red. Also consider R13 (skin) and R15 (Asian skin) if relevant.
4. TLCI (if applicable): mandatory for video, broadcast, and cinematographic applications.
5. TM-30-18 metrics (Rf, Rg, and graphs): the state of the art for a complete and modern evaluation.
6. Luminous efficacy (lm/W): to balance quality and energy efficiency, calculating the total required flux and consumption.
7. Consistency and stability: CCT and CRI should be consistent from batch to batch and stable over time and with variations in junction temperature and dimming.

Investing in high color quality LED sources means investing in the perception, comfort, and value of the illuminated space. Whether it's about selling more products, fully appreciating a work of art, transmitting a perfect image on television, or working in a healthy and pleasant environment, the depth and accuracy of the information provided by light are, ultimately, the measure of its professional quality. Ledpoint.it, with its selection of products with extremely high chromatic performance and specialized technical support, positions itself as the reference partner for professionals who do not intend to compromise on light quality.