On Frequency Modulated Signals in Reverberation Measurements

A point‐by‐point curve tracer, similar to that described by Norris and Andree but modified to permit variations in sound intensity to be measured by a substitution method, is employed to study the reverberant decay of pure tones and tones having various degrees of frequency modulation (so‐called “warble” tones). Straight lines (db vs. time) are fitted to the decay curves by an analytic process which is described. The deviations of individual points from the best straight line are computed and found to fall into systematic time patterns which are critically dependent upon frequency and upon the position of either microphone or loudspeaker in the sound chamber. Major characteristics of the time‐deviation pattern usually persist as the frequency modulation of the source is increased but the magnitude of the deviations may be reduced considerably. An approximate relation between the average deviation and the fractional frequency modulation (Δf/f) is given. In view of the frequency selectivity of absorption, twenty percent is recommended as a satisfactory, degree of warble for all frequencies. It is pointed out that large errors in the measured decay rate may arise from an ordered arrangement of the time‐deviation pattern. An evaluation of these errors in terms of the average deviation of the experimental points from linearity is given which is believed to be applicable generally to any sound decay measurements. It is concluded that the warble tone is distinctly useful for securing decay curves that are more nearly linear but that space effects may still exist which are capable of introducing serious errors.