British Kinematography (1950)

•July, 1950 tak : measurement of acoustics 19 find for this time /6o : 10-3 = e~k'% t60 = 2-3 -|r = i^T cA V or, since k = ^ptt6o = °16^ Fin m3 A in m2 of open window In the case of an exponential drop the reverberation time measured is a parameter which completely determines the character of the reverberation. In more complicated cases, for example when strong fluctuations are superposed on the exponential variation, there is, of course, no simple relation between the reverberation time and the character of the curve. In that case it is obvious that an average exponential curve can be drawn, disregarding the fluctuations, and a value of /6o can be found. This quantity is called the average reverberation time. It is true that the phenomenon is not thereby completely characterised, because the fluctuations have been disregarded, but it gives nevertheless the general character of the reverberation. If the reverberation consists of the superposition of different exponential functions each with its own constant k, perhaps combined with fluctuations, the obvious method is first to replace the curve by an average curve without fluctuations, and to analyse that curve into its purely exponential components. If one then determines teo for the separate components one obtains a number of reverberation times which together to some extent fix the character of the reverberation. 2— MEASUREMENT OF REVERBERATION There are various experimental methods of investigating the reverberation. The oldest is that of Sabine, by which only the reverberation time is determined. For that purpose the time was measured which elapsed between the moment at which the sound became inaudible. This method was later improved by the use of an entirely automatically working apparatus which determined electrically the time in which the intensity of the signal decreased by 60 db. We have seen, however, that the reverberation time depends very much upon different factors, and that it may have a very capricious variation, so that the sole determination of the reverberation time cannot give in this way a picture of the true variation of intensity. At the present time use is made of recording meters which have the advantage that the exact behaviour of the reverberation is recorded, making it possible to draw an average curve and to determine from it one or more average reverberation times. Moreover, from the curve obtained conclusions may be drawn about the acoustic properties of the room being studied. The principle of these recording meters is very simple. After being amplified, the voltage of the microphone is connected to the ends of a potentiometer which is then automatically regulated so that the voltage taken off" remains constant. The mechanical regulator at the same time drives a stylus which records the position of the potentiometer contact on a roll of paper.3 Requirements of Reverberation Meter The necessary high velocity of recording in connection with the unavoidable mechanical inertia, makes the whole mechanical part of this apparatus extremely delicate and difficult to transport. We have therefore developed a method which is not subject to these objections. This method has been named the method of exponential amplification. We will consider once again the *ase where the sound pressure varies purely exponentially so that : o * P = p0 e 2j-* sin 27r 57