What should recorded music sound like?  Should it strive to be the most accurate reproduction of a performance, as Edison wanted?  Or should the recording enhance and improve?  Or does it need to derive from any real performance at all?

Audiences of the 1910s were unable to discern between live singers and on-stage phonographs during “tone test” demonstrations.  Obsessive audiophiles of the 1950s were enamored of high fidelity reproductions of concert hall performances or dynamic, real-world recordings of thunderstorms and trains.  But most listeners today are accustomed to compressed audio, pitch-corrected vocals, sampled instruments, non-fluctuating tempos, and songs digitally assembled from countless tracks, parts, chords, and even individual notes.  Not all of these developments are necessarily bad.  The democratization of recording and the creative flexibility enabled by digital manipulation and editing are commendable.  Fewer live musicians playing physical instruments is perhaps not.

In Perfecting Sound Forever: An Aural History of Recorded Music, author Greg Milner gives us a fascinating chronicle of numerous debates and rivalries (e.g., cylinder vs. disc, acoustic vs. electrical, 45 rpm singles vs. 33 1/3 rpm LPs, analog vs. digital, synthesizers vs. samplers, high-fidelity vs. compressed convenience, etc.) in the history of sound recording.  As many worry about the aesthetic consequences of technological advancement, many of these discussions continue today.