Effect of Sample Heterogeneity on near Infrared Meat Analysis: The Use of the RMS Statistic

Ground and emulsified samples from Iberian pig hams were analysed by reflectance and interactance reflectance mode. Spectral errors due to intra-sample variations were calculated for both analysis modes. The spectral errors were calculated by means of the STD RMS statistic included on the ISI software. The results obtained show that a mean STD RMS value as low as 4200, could be obtained for paired subsamples of the same sample and that an STD limit of 4374 could be fixed at the instrument set-up program in order to ensure that a representative spectrum has been obtained from two subsamples readings of the same sample. That procedure avoids the need to take numerous subsamples, as is traditional in NIR/NIT meat analysis. The results also show that the spectral repeatability using fiber optic is worse than for spinning cups and it has been concluded that effort should be made to avoid moisture variations during scanning in order to improve spectral repeatability.