RECENT ADVANCES IN WAX RECORDING 169 



Moreover, it is cheap and durable, and reproduction of the sounds 

 from this record calls for no fine adjustments or intricate apparatus 

 as has been long evidenced by the broad use of the ordinary phono- 

 graph. 



The major part of the extraneous or "surface" noise found with 

 this method of reproduction comes from the material of the finished 

 record. Recent progress has been made in reducing this noise. As a 

 result of this, together with refinement in the plating processes, 

 records used with Western Electric Company theater equipment 

 during the last two years have shown a reduction of 3 to 6 TU in 

 "surface" noise. This corresponds to eliminating 50 per cent to 

 75 per cent of that previously present. It is not necessary to reduce 

 the level of "surface" noise to the zero point but merely to the 

 threshold of audibility under the conditions of minimum auditorium 

 noise which are of interest. This noise masks the surface. More- 

 over, it is not the absolute amplitude of the imperfection giving rise 

 to "surface" noise but the relative magnitude in comparison with 

 the useful sound amplitudes which counts. Thus, an effective re- 

 duction in "surface" could be made if we were willing to use larger 

 records or if we were willing to reduce the playing time of the present 

 records by increasing the spacing of the grooves and the amplitude at 

 which the grooves are cut. Any large reduction in "surface" noise 

 made by a reduction in the irregularities in the record material would 

 open the door to increasing the playing time of a record of given size. 

 There is no known absolute or fundamental reason why further 

 improvements in record materials may not be expected to reduce 

 further the amount of "surface" noise. Moreover, large advances 

 in pickup design open distinctly new possibilities as to reductions in 

 "surface." 



It has sometimes been thought that in order to reproduce high fre- 

 quencies properly, the linear record speed would have to be increased 

 or the size of the needle point reduced. At present the diameter of 

 the bearing portion of a representative needle is about .003 in. whereas, 

 as mentioned before, the half wave-length for a 5,000 cycle wave is 

 .0014 in. The factor determining whether a needle will follow the 

 undulation of the groove is not any consideration of the relative 

 diameter of the needle point and the undulation of the groove but 

 rather the radius of curvature of the needle and the bend of the groove. 

 As indicated before, the amplitude at 5,000 cycles would be only 

 about .0001 in. if sounds of that frequency were as intense as those of 

 lower frequencies (.002 in. at 250 cycles). As a matter of fact, sounds 

 of 5,000 cycles or more in speech or music are characterized by lower 

 12 



