Energy efficiency and colour fidelity in illuminating engineering
We have proposed new method for colour rendering assessment. In this approach, the colour shift associated with a test illuminant is evaluated for one thousand, or more,
representative reflectance spectra that span the full multi-dimensional range of possible spectral distributions and colours. This broad sampling eliminates the intrinsic selection bias of the
CRI calculation and its variants and could be used to develop a more objective standard for a colour quality score. Such a standard could be based on the fraction of the test spectra that undergo
colour shifts that are smaller than a just noticeable difference, or an agreed upon multiple of it. Since the concept of just noticeable differences in colour has already been reproducibly
quantified, it is hoped that this approach may be widely acceptable.