CRI/Ra显色指数全面指南

The Color Rendering Index (CRI), commonly denoted as Ra, is the internationally standardized metric that quantifies how accurately a light source reveals the colors of objects compared to a reference illuminant (incandescent light or natural daylight) of the same correlated color temperature.

\n\nThe Color Rendering Index (CRI), commonly denoted as Ra, is the internationally standardized metric that quantifies how accurately a light source reveals the colors of objects compared to a reference illuminant (incandescent light or natural daylight) of the same correlated color temperature. First established by the CIE in 1965 and most recently updated in CIE 13.3-1995, CRI remains the most widely used (CIE 13.3-1995, GB 50034-2013) color rendering specification in lighting procurement globally. This article provides a comprehensive technical explanation of CRI calculation, the significance of individual R-values (particularly R9), the limitations of Ra, and the newer TM-30-18 framework, with specific references to GB 50034-2013, EN 12464-1, and IES standards.\n\nThe importance of accurate color rendering extends across virtually every lighting application. In retail environments, studies have demonstrated that improving CRI from Ra 80 to Ra 90 increases perceived product appeal per CIE 13.3-1995 test methods by 15–30 % in controlled consumer surveys. In healthcare settings, accurate color discrimination under high-CRI lighting (Ra ≥ 90) supports clinical diagnosis of skin conditions, wound assessment, and medication identification. In museums and galleries, the CRI of exhibition lighting directly affects the visitor's ability to perceive subtle color variations in artworks — a difference of 5 points in Ra can significantly alter the perceived depth and texture of oil paintings under controlled viewing conditions.\n\nDespite its widespread adoption, CRI has well-documented limitations that lighting professionals must understand. The metric was originally developed for fluorescent and incandescent sources with relatively smooth spectral power distributions and has known shortcomings when applied to narrow-band LED sources. This has driven the development of complementary metrics including the IES TM-30-18 Fidelity Index (Rf), the Gamut Index (Rg), and the Color Quality Scale (CQS), which are increasingly referenced in high-end lighting specifications alongside traditional Ra values.\n\nHow CRI (Ra) Is Calculated: The CIE Test Color Method\n\nThe CRI calculation under CIE 13.3-1995 involves measuring the spectral power distribution (SPD) of the test light source and computing the resulting chromaticity shifts of 14 standard test color samples (TCS). The first 8 samples (R1–R8) are pastel tones of moderate saturation, used to calculate the general CRI index Ra. A reference illuminant is selected based on the CCT of the test source: a Planckian radiator for CCT below 5000 K, and a CIE daylight illuminant for CCT at or above 5000 K.\n\nThe eight test color samples used for Ra calculation are:\n\nR1: Light greyish pink (Munsell 7.5R 6/4) — moderate desaturated red\n\nR2: Dark greyish yellow (Munsell 5Y 6/4) — moderate desaturated yellow\n\nR3: Strong yellow-green (Munsell 5GY 6/8) — saturated yellow-green\n\nR4: Moderate yellowish green (Munsell 2.5G 6/6) — moderate yellowish green\n\nR5: Light bluish green (Munsell 10BG 6/4) — moderate bluish green\n\nR6: Light blue (Munsell 7.5PB 6/4) — moderate blue\n\nR7: Light violet (Munsell 2.5P 6/4) — moderate violet\n\nR8: Light reddish purple (Munsell 10P 6/4) — moderate reddish purple\n\nEach individual R-value (R1–R14, and optionally R15) is computed as a special CRI value using the formula: Ri = 100 − 4.6 × ΔEi, where ΔEi is the color difference in the CIE 1964 U*V*W* uniform color space between the test source and the reference illuminant for the i-th test color sample. The general CRI, Ra, is the arithmetic mean of R1 through R8:\n\nRa = (R1 + R2 + R3 + R4 + R5 + R6 + R7 + R8) / 8\n\nA source with Ra = 100 indicates perfect color rendering identical to the reference illuminant. A source with Ra below 50 produces heavily distorted colors, which is common with low-pressure sodium lamps (Ra ≈ 0–20) or some monochromatic LEDs.\n\nCRI Grading: Application-Specific Requirements\n\nThe required CRI varies dramatically by application. GB 50034-2013 specifies minimum Ra values for different building types, and EN 12464-1 provides equivalent guidance for European markets. The table below summarizes the standard CRI tiers and their typical applications.\n\nCRI (Ra) Range\nGrade\nTypical Applications\nGB 50034 Requirement\nEN 12464-1 Requirement\n\nRa < 70\nPoor\nUtility lighting, parking garages, security (non-critical)\nNot permitted for occupied indoor spaces\nNot permitted for indoor work areas\n\nRa 70–79\nBasic\nWarehouses, corridors, industrial storage areas\nRa ≥ 60 (Table 5.6.1, industrial)\nRa ≥ 60 (industrial)\n\nRa 80–89\nGood\nGeneral office lighting, classrooms, retail, hospitality\nRa ≥ 80 (Table 5.1.1, office)\nRa ≥ 80 (workplace interior)\n\nRa 90–94\nExcellent\nHigh-end retail, art galleries, printing, textile inspection\nRa ≥ 90 recommended (Table 5.4.1, museum)\nRa ≥ 90 (color-critical tasks)\n\nRa ≥ 95\nPremium/Museum\nMuseum conservation, medical examination, film/TV production\nRa ≥ 95 (Table 5.5.1, operating room)\nRa ≥ 95 (medical, color critical)\n\nAs of 2026, the Chinese lighting market has seen a notable shift: approximately 35 % of new commercial LED installations specify Ra ≥ 90, up from 12 % in 2020. This trend is driven by increased awareness of color quality in retail and hospitality environments and the availability of high-CRI LED chips at reduced cost premiums (approximately 15–25 % higher than Ra ≥ 80 equivalents at the chip level).\n\nR9: The Saturated Red Value and Why It Matters\n\nR9 is the special CRI value for test color sample #9 (TCS-09), which is saturated red (Munsell 4.5R 4/13, strong red). This is one of the six saturated test color samples (R9–R14) that are not included in the Ra average but are reported separately. R9 is critically important because:\n\nHuman skin tones contain significant red components; low R9 makes skin appear sallow or jaundiced.\n\nRed fruits, vegetables, and meat appear brownish or dull under sources with R9 ≤ 0.\n\nMedical applications: Accurate red rendering is essential for examining tissue, wounds, and rashes.\n\nFilm and photography: Low R9 causes poor color grading and requires extensive post-processing correction.\n\nMany standard white LEDs have inherently weak red emission because the blue LED pump and yellow phosphor combination produces a spectral "dip" in the 620–650 nm region. Typical mid-power LED packages achieve R9 values of 0 to +30. High-CRI LED chips using broad-spectrum phosphors or multi-phosphor blends can achieve R9 ≥ 90 at the expense of 5–10 % luminous efficacy reduction.\n\nProduct Category\nTypical Ra\nTypical R9\nEfficacy Impact\n\nStandard commercial LED downlight\n80–82\n0 to +5\nBaseline (120–140 lm/W)\n\nGood color quality LED panel\n85–88\n+10 to +30\n−3 to −5 % (115–130 lm/W)\n\nHigh-CRI LED module (Ra ≥ 90)\n90–93\n+50 to +70\n−5 to −10 % (105–125 lm/W)\n\nFull-spectrum LED (Ra ≥ 95, R9 ≥ 90)\n95–98\n+90 to +98\n−10 to −18 % (95–115 lm/W)\n\nMuseum-grade LED\n97–99\n+95 to +99\n−15 to −25 % (80–105 lm/W)\n\nBeyond Ra: TM-30-18, Extended CRI (R1–R15), and GAI\n\nWhile Ra has been the industry standard for decades, it has well-documented limitations:\n\nRa uses only 8 pastel test color samples, which are not representative of saturated colors (the Achilles' heel demonstrated by the R9 gap).\n\nRa averages R1–R8, so a source with poor R9 can still achieve Ra ≥ 80 as long as R1–R8 perform adequately.\n\nRa does not measure gamut area, so it cannot distinguish between a source with high fidelity but small gamut (which looks "washed out") and one with large gamut (which looks "vivid").\n\nThe IES TM-30-18 standard, introduced in 2015 and updated in 2018, addresses these limitations with two primary metrics:\n\nRf (Fidelity Index): The average color fidelity for 99 color evaluation samples (CES), providing a more comprehensive measure of color rendering than Ra. Rf typically correlates with Ra but can differ by up to 8 points, especially for sources with irregular SPDs.\n\nRg (Gamut Index): The average color gamut, where Rg = 100 indicates identical gamut to the reference, Rg > 100 indicates increased saturation (vivid), and Rg < 100 indicates reduced saturation (washed out).\n\nThe extended CRI system (R1–R15) includes R9–R14 (saturated colors: red, yellow, green, blue, light skin, and leaf green) and R15 (Asian skin tone, Munsell 5YR 8/4). While R15 is not part of Ra, it is frequently reported for lighting products targeting Asian markets, as it reflects skin tone rendering accuracy for light-to-medium skin tones more directly than R1–R8.\n\nMeasuring CRI: Equipment and Standards\n\nCRI measurement requires a spectral measurement device. The standard protocol follows IES LM-79-19 and CIE 13.3-1995:\n\nDevice: Array spectroradiometer (e.g., Instrument Systems CAS 140D, Konica Minolta CL-500A) or scanning spectroradiometer with wavelength accuracy ≤ 0.5 nm.\n\nSpectral range: 380 nm to 780 nm minimum; 350–830 nm preferred for extended calculations.\n\nWavelength resolution: ≤ 5 nm FWHM (full width at half maximum).\n\nIntegrating sphere: 1.0 m, 1.65 m, or 2.0 m diameter per LM-79, coated with high-reflectivity barium sulfate or Spectralon.\n\nCalibration: Traceable to NIST or NIM standard lamps with known spectral radiance.\n\nMeasurement uncertainty: For Ra, typical expanded uncertainty (k=2) is ±3 points for Ra ≥ 80 and ±5 points for Ra < 80.\n\nCRI is an inherent property of the light source and is independent of the luminaire's optical efficiency, beam angle, or power factor. However, if the luminaire uses mixing optics or