"Understanding the evolution of mainstream music is of high interest for the music production industry. In this context, we argue that a MIR perspective may be used to highlight, in particular, relations between dynamics and various properties of mainstream music. We illustrate this claim with two results obtained from a diachronic analysis performed on 7200 tracks released between 1967 and 2014. This analysis suggests that 1) the so-called “loudness war” has peaked in 2007, and 2) its influence has been important enough to override the impact of genre on dynamics. In other words, dynamics in mainstream music are primarily related to a track’s year of release, rather than to its genre."(I assume "MIR" stands for Music Information Retrieval.)
In their conclusion, they say:
"We have shown that the loudness war has peaked in 2007, and that a return to pre-loudness war dynamics may be reached in about ten years. As an exception, macrodynamics, which have not been significantly influenced by the loudness war, appear to increase since the loudness war’s peak, and are currently reaching very high values."The article includes plots showing the evolution of RMS power, EBU3341 integrated loudness, crest factor, PRRC, HLSD and EBU3342 Loudness Range over the years. It's great to see such a thorough analysis on the topic.