Microphone at 2 meters distance and a height of 1 meter.
NOTE: Although these are 512 point FFT plots, there is NO WINDOWING
of the impulse data. The low-frequency response shown on these plots
contains measurement artifacts - which is precisely the reason that we
window the data!
The three plots/averages were taken one after the other, under as
similar conditions as could be controlled. Clearly, averaging multiple MLS
stimuli affects the measurement of the low frequencies. One explanation for
this, based upon experience, is that the amount of low frequency noise
contamination in a typical home is large. Averaging tends to reduce the
effects of such contamination.
MLS: 1, 32 or 64 averages, NO windowing.
1/12 octave smoothing - Amplitude Axis @ 6dB per division.
Measurement Issues - Reflections in Impulse Responses.
Measurement Issues - Lies, Damned Lies & Statistics.
Acoustic Line Source Research - Table of Contents.