The statistical analysis was performed in SAS 9.2. The model validation was primarily done by evaluating the residuals (and the so-called studentized residuals) obtained after fitting the models for all three parameters (L*, a*, and b*). The plots used in the validation showed very similar patterns for the three parameters, and they will therefore be described In Fig. 4 the different kinds of samples considered in this experiment are illustrated. This illustration indicates that the colors returned by the two instruments are not the same. The ΔEab⁎ values of the average color values are illustrated in the top left plot in Fig. 5. A clear difference in magnitude of ΔEab⁎ depending on the processing level of the product is observed. For the fresh products the ΔEab⁎ values are all above five, whereas for the processed types, the values are all under five. This clear differentiation between the fresh/raw and processed samples can be related to the threshold for human detection of color difference. Larraín et al. (2008) states that no exact threshold has been established for meat color difference detection, but that a possible value could be around 2–6. This explains the visual differences observed in Fig. 4.