According to strict laboratory procedures, we should have made exhaustive tests and calculations and six months later come up with the correct answer. What happened was that we got together with Warner’s chief projectionist and asked him how fast they ran the {silent} film in theaters. He told us it went at eighty to ninety feet per minute in the best first-run houses, and in small ones anything from one hundred feet up, according to how many shows they wanted to get in during a day. After a little thought, we settled on ninety feet a minute as a reasonable compromise. We had our disks processed by the commercial record companies, and the largest diameter they could handle was about seventeen inches. With a record of that size, the optimum speed to get the ten minutes of recording time we needed was somewhere around 35 revolutions a minute. We standardized at 33-1/3, because that happened to fit the best with the gearing arrangement our engineers were working out for coupling the turntable to the picture machine. Western Electric Engineer Stanley Watkins (1961) Scott Eyman THE SPEED OF SOUND: Hollywood and the Talkie Revolution 1926-1930. p. 112.33-1/3 was used for early sound film into the early 1930s where sound on film finally became the standard. The 16” 33-1/3 record continued to be used for radio transcription, and for syndicated radio shows. The 33-1/3 twelve inch LP was introduced for consumers by Columbia in 1948.
RCA developed the 7” 45 in 1949. They also used the larger 1-1/2 center hole to make it easier for DJs to handle the record. So why 45?
The clarity of a 7" record turning at 33-1/3 isn't very good even with the new microgroove stylus. How did they arrive at 45? This is my theory. RCA had a number of transcription turntables that ran 78 and 33-1/3. The drive motors were geared 1:23 from 1800 RPM to 78.26 RPM. Just under the platter was a planetary ball bearing. The inner race rotated at 78.26 RPM. The five balls were in a keeper that drove the platter. When the outer race was locked to the platter, the balls and outer race were locked making the whole bearing turned at 78 RPM. When the outer race was locked stationary, the balls and hence the platter turned at 33-1/3 RPM. The RCA 70-C and 70-D series were in large production, and used in most radio stations. This bearing was very unusual as the inner race was machined to 1-79/128”, the outer race to 2-23/128” and used standard 9/16” balls. Not exactly an off the shelf size. Since planetary gearing works on subtraction, if you were to rotate the outer race at 78, and hold the inner race stationary, the balls would turn at 45 RPM, well actually 44.92753623 RPM. Close enough. This also allowed the standard 1800 RPM motor geared at 1:40 to produce 45 RPM. With the smaller microgroove stylus’s, the groove velocity with the smaller diameter at 45 RPM still could be as clear as the larger standard stylus’s at 78 RPM. 16-2/3 RPM and 8-1/3 RPM are just halving of the 33-1/3 RPM. These were used for Talking Books for the blind. Since these were voice only, running the tape at double or quadruple speed accomplished the half 16-2/3 or quarter 8-1/3 speeds for playback. 16-2/3 was also used for Seeburg Basic Music Library, a competitor to Muzak in stores, and the Highway Hi Fi car record player.