38 SECTIONAL ADDRESSES 



levels. If the network is omitted, the meter measures intensity levels. 

 Objective meters should be further designed to simulate the salient 

 characteristics of the ear in dealing with either steady or impulsive sounds. 

 For example, the ear does not record full strength until a steady sound has 

 persisted for about one-fifth of a second, and in the case of pulsating 

 sounds, the inertia of the instrument has to be of the right order to give 

 readings corresponding to average aural appraisement. 



The practice has developed in certain countries of constructing 

 objective meters to arbitrary specifications (which include quadratic 

 rectification and certain instrumental requirements), the meters being 

 ostensibly designed to deal with pure tones and certain common types of 

 sound. The readings, which are sometimes referred to as ' sound levels,' 

 are not claimed as conforming necessarily to subjective equivalent loud- 

 nesses. Experimental calibrations of such meters show, however, that 

 while they may give results in reasonable agreement with subjective 

 measurements in the case of steady pure tones or other sounds, they 

 afford readings in the case of particular types of intermittent noise (such 

 as certain machinery or traffic noises) which are considerably lower than 

 equivalent loudnesses as measured by the fundamental method. It was 

 agreed at the international acoustical conference in Paris last July that it 

 was undesirable that a meter should be used for measuring the equivalent 

 loudness of sounds unless it has been shown, by calibration in an accepted 

 standardising laboratory, to give results, for the particular sounds in ques- 

 tion, which are in reasonable accordance with the subjective scale of phons. 



The difficulty of dealing with intermittent sounds can, it appears, be 

 met by replacing the quadratic rectifier by one of the leaking peak variety. 

 Portable objective meters of this type have been designed by Davis at 

 the National Physical Laboratory and independently in Switzerland. 

 The former instrument has recently emerged successfully from a pro- 

 longed series of tests on the equivalent loudness of a wide variety of noises, 

 including continuous, warbling, and impulsive short-duration sounds 

 (recurring at rates of from 12 to 50 per second), as well as the noises of 

 motor cars, motor cycles and motor horns. For moderate or loud noises, 

 the measurements were found to agree more closely with the average aural 

 appraisements of a group of observers than did those of any one of the 

 observers. The leaking peak rectifier enables the rates of integration 

 and decay to be adjusted empirically so that the meter simulates the 

 response of the ear in giving higher results for rapidly recurring impulses 

 than for slowly recurring or single impulses. For example, the loudness 

 of a single impulse may be nearly 10 phons less than if 30 to 50 such 

 impulses occur per second. Experiments designed to formulate a 

 specification of the physical behaviour of the meter are not yet complete 

 and, at present, the meter is not designed for levels lower than about 

 65 phons. 



While, however, a * universal ' objective meter is not yet speci- 

 fiable, there can be little doubt that the future of sound or noise 

 measurement, from a practical point of view, lies with the objective meter. 

 Even in its present state of development, its practical convenience often 

 outweighs its limitations. It is free from personal bias, rapid in action 



