The basis of stellar population modeling was established around 40 years ago somehow optimized to the technical facilities and observational data available at that epoch. Since then, it has been used extensively in astronomy and there has been great improvements relating their associated ingredients in concordance with the development of more powerful computational and observational facilities. However, there has been no similar improvements in the understanding about what is actually modeling neither in improve the modeling itself to include the current technical advances to obtain more accurate result in the physical inferences obtained from them. In this talk I present some advances in the subject of stellar population modeling and how to take advantage of current facilities to obtain more robust and accurate inferences from stellar systems at different scales covering the continuum between fully resolved populations to fully unresolved ones in a unified framework.