Different measures of sensitivity of recipe colour to random and proportional dye concentration error. Part 3: Observed repeatability in regard to predicted sensitivity

A regularity of the predicted sensitivities to random and proportional dye concentration errors in regard to the position of target colour has been observed for the case of dyeing acrylic with basic dyes. The sensitivity of the recipe colour to random dye concentration errors is highest for light neutral target colours and is almost negligible for dark-shade recipes. Recipes for less saturated targets are slightly more sensitive than recipes for more saturated targets of equal lightness. The span of the sensitivities to weighing error of the recipes matching a given target varies with the position of the target in colour space. By contrast, the sensitivity of recipes to dye strength error is the highest at medium to low lightness for neutral and near neutral target shades. The span of sensitivities of the recipes for any such particular target is broad, with some having low sensitivity to strength errors. Recipes for the target colours at the "lighter" part of the gamut border were the least sensitive to strength errors.