Page 17 - 2016Fall

Figure 2. Steps in the process from wax to published record. The flat wax “cake” is given a shiny surface, shipped to where it is recorded, and then shipped to the factory electrotypical department where a stamper for the records is prepared. Copyright © 1983 George Brock- Nannestad.
In the beginning, only the cornet, banjo, and tenor voice were really good for recording this way, but the record com- panies wanted a large repertoire and also a resemblance of realism to the listener. For this reason, they did thousands of experiments, condensing the results as suitable combina- tions of soundboxes, diaphragm material, recording horns (the funnels for collecting the sound), and distances from the recording horns, all of them trade secrets. The record companies had to balance a fine line between background noise and distortion.
The horns used for recording were invariably conical, where- as those for reproduction were mostly flared, although not truly exponential as taught by theory. The recording com- panies’ musical staff also rearranged the music to support string bass lines by deep brass instruments and specified so- called Stroh instruments for the string section (Stroh, 1900). Instead of a body, they only had a skeleton outline and were fitted with a diaphragm in a huge soundbox. A Stroh violin is shown in Figure 3. The diaphragm was activated by the bridge, and the soundbox was connected to a horn that was intended to point in the direction of the recording horn.
There was no meter to determine the recorded level of the sound, and the only reliable measurement was that of the durability of the manufactured record: after how many plays would the wear distortion be so bad as to be uncommercial? This was determined by playing a test record until it broke down. If the count was satisfactory, a whole batch of records would be manufactured. The number varied among compa- nies, but a minimum of 50 reproductions on the company’s own brands of reproducer (gramophones and phonographs) was a requirement. Some combinations of manufacturing procedures, materials, and reproducers could provide as many as 200 reproductions.
Figure 3. For recording purposes, J. A. Stroh invented this instrument with the outer dimensions of a violin but with a horn to direct the sound toward the recording horn. The bridge (b) rests on a transverse rocking lever (c) with knife edges (d-d). A lever (not visible) going sideways transmits the vibrations to the center of a large diaphragm in the soundbox (o). The big radiating horn was fitted to the sound- box. From Stroh (1900) patent drawing.
It was discovered that some of the wear was reactive (we would say a mismatch of impedances) and some of the resonances in the stylus-soundbox-horn combination were quite sharp. Thus, one company, the Victor Talking Ma- chine Company, recorded at 76 rpm and specified 78 rpm for reproduction. This effectively staggered the tuning and avoided exciting resonances with signals that were already strongly recorded due to resonances in the recording setup (Brock-Nannestad, 1997). All commercial sound recording developed by trial and error, for instance, a good-wearing groove profile was discovered after examination of innumer- able microscopic groove section photographs and relating them to the durability determined by the wear tests.
Fall 2016 | Acoustics Today | 15