The International Commission on Illumination (CIE) issued a Position Statement on Color Quality Metrics, its second edition, recommending the improvement of color rendering metrics for white light sources. The statement highlights the development of the CIE General Color Fidelity Index (Rf) in 2017 as a solution to technical issues with the older CIE General Color Rendering Index (Ra).
Initially, Rf was recommended for scientific use rather than industrial applications. However, after extensive use without contradictory evidence, the CIE now recommends that the lighting community adopt Rf to replace Ra in relevant regulations, specifications, and standards.
To ensure a smooth transition, the CIE suggests reporting both metrics in parallel until Rf becomes widely established. This approach allows for an orderly shift in industry practices while maintaining continuity.
The CIE statement acknowledges that a comprehensive understanding of color quality extends beyond fidelity alone. As such, the CIE recognizes the need for developing additional metrics to address color preference, indicating ongoing work in this area. IES TM-30 utilizes a combination of Rf for color fidelity and Rg for color preference.
The updated CIE position reflects the desire to improve color quality assessment in lighting, adapting to advancements in both light source technology and color science. The full CIE Position Statement provides more detailed information on these recommendations and the rationale behind the recommended transition from Ra to Rf.
Key Differences Between CRI (Ra) and CFI (Rf):
The Ra (CIE General Color Rendering Index) and Rf (CIE General Color Fidelity Index) are both metrics used to assess the color rendering quality of light sources, but they have several important differences:
- Number of color samples: Ra uses only 8 color samples, while Rf uses 99 color samples. This larger sample set in Rf allows for a more statistically representative and reliable metric. The 8 Ra samples are part of a set of 15 total color samples established by the CIE. The first 8 color samples, which are used to calculate Ra, are moderately saturated colors that are evenly distributed across the hue spectrum. While there are additional color samples (9-15) that can provide supplementary information about color rendering properties (i.e. R9), particularly for saturated colors and skin tones, these are not used in the calculation of the general Ra value. The Ra is specifically calculated as the average of the color rendering indices for the first 8 color samples.
- Color space: Ra calculations are based on the outdated CIE 1964 Uniform Color Space, whereas Rf uses the more advanced CAM02-UCS color space.
- Calculation method: Rf employs a different calculation method that includes a scaling factor to align its values with Ra for a set of reference light sources.
- Scope of application: While Ra has been widely used for rating and specifying products, Rf was recommended primarily for scientific use and has not considered a direct replacement for Ra in product specifications or regulatory requirements.
- Sensitivity to spectral power distribution: Rf appears to be more sensitive to narrow-band light sources, potentially providing a more accurate assessment of light sources specifically designed to achieve high Ravalues.
- Real-world representation: The 99 color samples used in Rf are chosen from real-world objects and categorized into groups such as nature, skin color, textiles, and paints, allowing for more application-specific evaluations.
Despite these differences, both metrics show high correlation with subjective color differences in many cases. However, Rf addresses some of the technical limitations of Ra and provides a more comprehensive assessment of color fidelity, particularly for LED light sources.
Image: CIE
You must be logged in to post a comment.