180 BELL SYSTEM TECHNICAL JOURNAL 



tically, we have to recognize the existence of ground noise in all 

 records and take precaution to minimize it. No matter how ex- 

 cellently we reproduce the fortissimo passages, our record is unsatis- 

 factory unless the ground noise is low enough for a wide volume range, 

 that is, a wide range in level between fortissimo and pianissimo. 

 Whether our negative sound record is made on negative or positive 

 emulsion, there is always the danger that in reproduction we shall 

 encounter variations in transmission from point to point due to local 

 variations in the celluloid base, to local action of the developing agent, 

 or to a developer excessively granular in action. The photoelectric 

 cell is able to recognize variations of 1/10 of 1 per cent, whereas the 

 eye ignores contrasts under 2 per cent. These local variations in 

 transmission, continued to the positive print, constitute the ground 

 noise. 



The remedy is, in part, to choose a developer as little granular in 

 its effect as possible. In part, to insist on machine development of 

 the sound film with thoroughly agitated developer. Further, to 

 carry the sound development to a high gamma; this obviates to a 

 large extent flow marks of the developer, and goes a long way to 

 escape local variations in the base by developing the negative striations 

 to be conspicuous in comparison. 



In 1924 we concluded that the optimum choice was positive emulsion 

 developed to unit gamma for both sound negative and sound print. 

 This is feasible for sound records separate from pictures, but a compro- 

 mise must be made for the combination of sound and picture in a 

 single positive print. Here the positive development required for a 

 satisfactory picture is always to a gamma far above unity. 



It is customary to develop picture negatives by inspection, having 

 in mind the uniform positive development to be undergone by the 

 prints from these negatives. The gamma of these positives need 

 never exceed 1.8; the sound negative then should be developed to 

 0.55. In order not to disturb the practice of the film laboratory, we 

 ask that the positive development be standardized and its gamma 

 ascertained, the reciprocal of this gamma then arranged for in the 

 standardized negative development. A negative gamma above 0.5, 

 together with the precautions of careful handling, permits the realiza- 

 tion of an adequate volume range. 



It is beyond the scope of this paper to discuss the details of manipu- 

 lation and of choice of developer, but I wish to acknowledge the 

 cooperation of Mr. J. W. Coffman in the solution of such problems. 

 The problem is the reduction of ground noise, and its seriousness is 

 not to be diminished by choosing a different recording method. 



In printing the sound negative, a uniform density for the print of 

 the unmodulated track is desired. The volume of reproduced sound 



