The colour shade quality control scenario is ubiquitous within the cosmetic hair colouring industry. Visual inspection is nowadays the go‐to standard for the evaluation of dye application outcomes, even although the increased availability of easily operated spectrophotometers has favoured the introduction of colorimeters in the manufacturing process. Human hairs, however, are very efficient scatterers, and this makes their instrumental measurements more difficult and less consistent. To assess the degree to which this intrinsic property compromises the reliability of spectrophotometric measurements, we tested a variety of samples of human hair locks (swatches), both dyed and undyed, and those of nylon fibre hair swatches. First, we analysed results from two measuring modalities made available by instrumentation: Specular Component Included and Specular Component Excluded. Then, to modify macroscopic spatial orientation, we arranged swatches so that hair strands would cross at varying angles. Measurements confirmed the expected dependencies, namely, the clear influence of the scattered component and of sample orientation on instrumental reading outputs. Having settled on a preferred measuring setup, we analysed measurement variance. Here, we briefly discuss the qualitative visual inspection of spectra, before switching to ΔE to show the unreliability, at least as far as lighter coloured dyes are concerned, of the classic target‐sample measurements which are employed to issue pass‐or‐fail verdicts. Despite quantitative measurements being an asset for quality control, we suggest that greater care should be placed in spectrophotometric readings when human hair is involved, as opposed to more traditional fields of application.

中文翻译：

---

头发光谱反射率和颜色测量的一致性问题

在化妆品染发行业中，色泽质量控制方案无处不在。如今，目视检查已成为评估染料应用效果的首选标准，即使易于操作的分光光度计的可用性不断提高，也有利于在生产过程中引入色度计。然而，人的头发是非常有效的散射体，这使得它们的仪器测量更加困难且不一致。为了评估这种内在性能损害分光光度法测量的可靠性的程度，我们测试了各种染色和未染色的人发lock（色板）样品以及尼龙纤维发色样品。首先，我们分析了仪器提供的两种测量方式的结果：包括镜面反射分量和排除镜面反射分量。然后，为了修改宏观空间方向，我们安排了色板，以使发束以不同角度交叉。测量结果证实了预期的依赖性，即分散成分和样品方向对仪器读数输出的明显影响。确定了首选的测量设置后，我们分析了测量方差。在此，我们简要讨论光谱的定性视觉检查，然后切换到ΔE以显示至少对于涉及较浅色染料而言不可靠的经典目标样品测量，该测量用于发布通过或不通过的结论。尽管定量测量是质量控制的资产，