Three Minutes, Because of a Disc

A ten-inch disc turning at seventy-eight revolutions per minute holds about three minutes of music on a side.

That is a fact about shellac and groove spacing, and for roughly fifty years it was also the length of a popular song. Arrangements were cut to fit. Verses were dropped, solos shortened, introductions trimmed. Bands rehearsed with a stopwatch, and Duke Ellington's longer pieces had to be split across two sides, which is why some of them have a structural seam at the point where the record had to be turned over.

By the time the constraint disappeared, the length had become the convention, and it largely remains one. Musical form settling around a manufacturing limit is not unusual (the fresco's giornata and the sonnet's fourteen lines are the same kind of fact) but it is startling how completely the arbitrary becomes the natural.

The deeper change was that a performance became an object.

Before recording, music was an event. It happened, and then it was over, and what survived was notation, memory and description. After 1900 a performance was a thing you could buy, replay, slow down, and study bar by bar.

Musicians immediately learned from it. Players lifted solos note for note off discs, and vocabulary that would have taken a generation to travel by ear and apprenticeship crossed a continent in a year. It also cut the other way: recording preserved regional and rural music that would otherwise have vanished without trace, and simultaneously eroded regional difference by giving every young player the same models.

Both effects are real, which makes the standard argument about whether recording helped or harmed folk traditions unanswerable in that form.

Then the technology changed who could be recorded at all.

Acoustic recording worked by shouting into a horn that drove a cutting stylus directly. It rewarded loud, focused sources: brass, banjo, trained operatic voices. It buried string basses, quiet singing and anything with a diffuse sound. Ensembles rearranged themselves physically around the horn, and instruments were substituted for ones that recorded better.

Electrical recording with a microphone arrived in 1925 and inverted all of it. A singer could now be quiet, close and conversational and still be heard, which is why crooning exists as a style. Bing Crosby's entire manner is a response to a transducer, and the intimate popular vocal that has dominated ever since begins there.

Finally, tape, and the end of the take.

Before editing, a take was a performance: what you hear is what happened, and a mistake in bar ninety meant doing it all again. Magnetic tape from the late 1940s allowed splicing, and the released performance need never have occurred as a whole.

Classical recording became a matter of assembling an idealised reading from many passes. Popular recording went further and dispensed with the idea that a record documents an event at all, which is the subject of the next chapter.

For the listener, the useful habit is to ask what a recording is claiming to be. A document of something that happened in a room, an idealised version of something that could happen, or an object that could not be performed at all. All three are legitimate. They are different art forms sharing a delivery format, and a great deal of argument about music is really a disagreement about which one is on the table.