Development of the next-generation space flight vehicles has prompted a renewed focus on rocket sound source characterization and near-field propagation modeling. Improved measurements of the sound near the rocket plume are critical for direct determination of the acoustical environment both in the near and far-fields. These measurements are also crucial inputs to empirical models and utilized in validating computational aeroacoustics models. NASA's SP-8072 acoustic load prediction model [K. M. Eldred, NASA SP-8072 (1971)] is a widely used empirical model for predicting liftoff acoustics. The model implements a Distributed Source Method, which predicts the loading as the sum of the radiated acoustic pressure field from each source distributed along the plume. In this paper, the resultant sound distributions are compared among the original core length definition in SP-8072, a modification proposed by Varnier [J. Varnier, AIAA Journal, Vol. 39, No. 10, (2001)], and four full-scale solid rocket motor tests with measurements covering a spatial region of 9 to 56 nozzle diameters downstream of the nozzle. The results of the comparisons support the original Eldred core length definition on an overall level basis. However, analysis of the measured spectra unveil shortcomings within both definitions.