A loudspeaker is untimately just a balance of compromises: ask any experienced listener, engineer, or reviewer! This is because the process of transducing an electrical signal into an acoustic one is, despite over a hundred years of development, still not a perfect science. Making a device that can sit in a living room and reproduce everything from a full-scale orchestra to the human voice, and all points in between with equal accuracy, is a herculean task. Which was perhaps why in the late 1960s, Yamaha made the NS20, a speaker with a very large rectangular driver on the back that was designed to emulate a cello by approximating its size and shape. Neat idea, but what chance did that have of reproducing an organ or bass guitar without screwing up.
The sheer difficulty of the job that a loudspeaker has to do has led to a diverse array of solutions. The first loudspeakers were simple horns, which still have a following today because nothing else can deliver dynamics and speed in quite the same way. Next came the infinite baffle with a conical driver, the simplest and least expensive solution, and the one which ultimately has proved most enduring.
The loudspeaker has come a long way since its introduction in the late 19th century, albeit not perhaps as far as the early pioneers of the technology might have imagined given the pace of change at the time. The moving coil transducer was patented in 1874, and the first conical diaphragm in 1901. Rice and Kellogg came up with the principle of the direct radiator in 1925, and Kellogg filed a patent for an electrostatic speaker in 1929. Even the first two-way loudspeaker appeared in 1931, and the acoustic suspension system, cornerstone of every small box loudspeaker ever since, dates back to 1954. Engineers now largely refine these principles than trying to start afresh, in an attempt to iron out their shortcomings and improve bandwidth, linearity, and power handling.