Electrical sound recording, which relies on microphones and amplifiers, became the industry standard around 1925. From Edison’s invention of the phonograph in 1877 to the mid-1920s, sound was only recorded acoustically, the sonic vibrations transmitting themselves directly from the horn to the diaphragm and the needle, unamplified. The recording process was uncomfortable for the players and the resulting tracks on master disks or cylinders omitted or distorted much of the audio input.
Still that was better than nothing. Speech, unaccompanied singing, and the cello were sometimes recorded quite decently thanks to their modest frequency range when the speaker or performer refrained from drastic volume shifts. Uniformly loud was the best. Even so, there were several attempts to record full-size orchestras and choirs performing large-scale works. Played on the typical affordable gramophone of the 1910s, these records certainly left much to be desired. Sound restoration engineers such as Ward Marston and Mark Obert-Thorn have done great work bringing old tracks back to life, exposing a wealth of sonic data hidden within them.
In other words, there is hope for even the least intelligible ancient recordings, and the question lingers: why did not they record more? It’s great that Mahler left a few Welte-Mignon rolls but a sound recording of his conducting, however flawed, would have added to our knowledge of his musical thinking. There is a miraculously extant recording of Tchaikovsky, Anton Rubinstein, Yelizaveta Lavrovskaya and their friends testing a cutting-edge phonograph and enjoying themselves. It was 1890. A great composer, a great pianist (and major composer) and one of the finest mezzos of her time shouting into a big brass horn- and that was it. No attempt to record any music. Tchaikovsky would die in 1893 and Rubinstein in 1894.
(In contrast, Brahms (who died in 1896) played a little for Edison’s agent in 1889. Those cylinders were once thought useless because of the overwhelming noise, but that was a premature opinion.)
Are we making the same mistake, perhaps, as the skeptics of the phonograph era? Are we under-capturing some important phenomenon for posterity on the pretext that the technology for that is still embryonic? It’s probably worth thinking in that direction, just in case.