A.— MATHEMATICAL AND PHYSICAL SCIENCES 37 



simpler techniques have accordingly been evolved which make portable 

 noise meters feasible for field conditions. There are two main types of 

 meter on the market, namely subjective meters, which were first developed 

 by Barkhausen, and objective meters utilising a microphone and amplifier 

 system. At the National Physical Laboratory, the calibration of such 

 meters in phons is effected in lagged chambers by means of sounds of 

 known phon values, for example, pure tones of various frequencies, and 

 gramophone, synthetic or other reproductions of a variety of noises such 

 as the meter may be called upon to measure in practice. The British 

 Standards Institution is at present engaged in developing specifications 

 of noise meters designed to conform to the B.S. scale of phons, and, in 

 the meantime, purchasers of noise meters would do well to demand a 

 calibration certificate by an accepted testing authority. 



(i) Subjective noise meters. — Subjective noise meters depend on the 

 equality matching of the loudness of a noise, as heard by the ear, with a 

 reference tone (usually a pure tone of specified frequency and of graduated 

 intensity) as heard in a telephone earpiece held tightly against one ear. 

 The reference tone may be produced by an electric buzzer, valve oscillator 

 or other means, various frequencies (e.g. 1,000 or 800 cycles per sec.) or 

 mixed tones being used in commercial instruments. The usual 

 Barkhausen technique involves simultaneous listening of the noise and 

 the reference tone, but experience indicates that inconsistencies which 

 are found to arise in aural measurements under such conditions are largely 

 resolved when the two sounds are heard alternately for periods of not less 

 than a second. Such a technique has been facilitated by Churcher in a 

 subjective meter employing two earpieces, so that either the noise or the 

 reference tone can be heard in turn using both ears simultaneously. 



Subjective meters are useful for certain purposes, but the aural judging 

 of equality of loudness of the reference tone and of a noise very diff'erent 

 in character is not always easy. Observers are found to differ widely 

 in their judgments and the same individual is not always consistent. 

 The accuracy of appraisement of a single observer is normally low (say 

 5 phons) and a team of trained observers is essential for higher precision 

 (say 2 phons). In the case of unexpected or single impulsive sounds, the 

 aural assessment presents great difficulties to the average observer. 



(ii) Objective noise meters. — The problem of the designing of objective 

 sound and noise meters, which has received much attention both in this 

 country and abroad, is of considerable complexity. The ideal aimed at 

 by objective meters is to be able to measure every type of sound and noise 

 on the subjective scale of phons, that is to simulate the selectiveness and 

 response of the average ear in all circumstances. 



The various objective meters on the market all consist essentially of a 

 pressure microphone connected to an amplifier provided with calibrated 

 control, followed by some type of rectifier and an output indicator. As 

 a first essential, objective meters are constructed to give the same reading 

 for a range of steady pure tones which sound equally loud to the ear 

 whatever the frequency. This is achieved by introducing into the 

 amplifier electrical networks designed to modify the shape of the frequency 

 characteristic so that it imitates the ear sensitivity at selected loudness 



