The ability to estimate 24 h pH values on intact meat from NIR spectra
on homogenized meat was poor (R2 = 0.73; RMSECV = 0.09; RPD =
1.14), which is in accordance with that found by Prieto et al. (2008)
and De Marchi et al. (2013) in beef samples. The failure of NIRS to
estimate pH value in this study could be due to the sample preparation
(intact for referencemethod vs. ground for NIR spectra), since scanning
the samples after grinding could reduce the precision of pH estimations
due to a lack of information about themuscle structure i.e. light scattering
properties in intact muscle tissue (Prieto et al., 2008). Additionally,
when measuring pH on intact meat the pH values are mainly from
lean, since pH measurements are taken in those homogenous parts of
the muscle without apparent marbling. However, in the ground
samples, pH values are from both lean and fat, so pH values from the
same muscle could be different depending on the sample preparation.
In this sense, ElMasry, Sun, and Allen (2012) showed howthe pH values
vary from sample to sample and from location to location in the same
muscle depending on the marbling.