Color Quality Scale

• The color space used in CIE R_a is outdated and nonuniform, and CQS uses CIELAB as a replacement.
• The Von Kries chromatic adaptation transform used by R_a does not perform as well as other available models. CQS uses CMCCAT2000.
• CIE R_a is based on desaturated samples, and a lamp's performance in rendering these samples fideliously is not necessarily linked to how it may perform with samples of higher saturation. CQS uses higher saturation samples.
• ‘Pure’ fidelity does not account for desired chromaticity changes. Increased saturation might be preferred. CQS does not penalise against increases in saturation.
• In CIE R_a the arithmetic mean is taken of the color differences for the individual samples. In CQS the individual results are combined through a root mean square instead, so that a small number of poorly rendered objects reflects with greater strength in the overall result.
• Negative values of CQS are made impossible due to their potential for consumer confusion.
• CCTs of lower than 2800K are penalised so that the CQS is more representative of their actual color rendering as opposed to their fidelity.