Accurate method for computing correlated color temperature

0
259

The article titled “Accurate method for computing correlated color temperature” by Li et al., published on June 27, 2016, addresses the problem of calculating the correlated color temperature (CCT) of a light source, proposing an accurate method based on the use of Newton’s method. In this work, the authors point out that previous available methods for calculating CCT use approximate objective functions, which limits the accuracy of their predictions. These limitations become evident when different unacceptable CCT values are predicted for light sources located on the same isotemperature line.

The approach proposed in the article is based on Newton’s method, which requires the first and second derivatives of the objective function. The authors explicitly provide these derivatives, following the current recommendations of the International Commission on Illumination (CIE) for the calculation of tristimulus values with 1 nm sums from 360 nm to 830 nm. This allows precise calculations of the CCT with errors less than 0.0012 K for light sources with CCT between 500 K and 10^6 K.

In the design and evaluation of their method, the authors conducted exhaustive tests that demonstrated the high accuracy of the results obtained using Newton’s method, along with an initial estimation of the CCT using Robertson’s method. The accuracy was such that for some light sources with CCT below 5000 K, the maximum error was less than 1.5 × 10^-7 K. However, it was observed that as the CCT increased, the maximum error also increased.

The results obtained were compared with other methods, such as the AR method (Robertson with 30 and 60 intervals), the JS method, among others, and it was shown that the proposed method surpasses the others in accuracy, although it also took approximately 10 times more computational time per sample. Specifically, it was demonstrated that the proposed method converges with no more than nine iterations within a tolerance of 0.001 K for CCTs ranging between 500 K and 10^6 K.

Furthermore, the CIE recommendations were highlighted, suggesting the use of 1 nm sums for the calculation of the TSVs. Although summing with larger intervals could accelerate convergence, the accuracy decreases, making 1 nm sums still the most accurate option.

The complete implementation of the method in MATLAB includes various stages, from the calculation of tristimulus values to the iteration in Newton’s method to adjust the CCT. Each step is defined in detail, ensuring accuracy in the color temperature optimization process.

The discussion in the article also mentions that, although the proposed method is superior, it strongly depends on a good initial estimation of the CCT. The authors suggest that this initial estimation be determined using the AR method due to its robust performance.

In conclusion, the article establishes that the proposed method offers a practically exact solution for calculating the CCT, validating its claims through extensive tests and comparisons with existing methods. The recommendations and detailed analyses presented provide clear guidance for the precise and efficient implementation of CCT in various industrial and scientific lighting applications.

Review on the publisher’s website