Measurement Issues - Comparison: MLS Averaging.

Microphone at 2 meters distance and a height of 1 meter.
MLS: 1, 32 or 64 averages, NO windowing.
1/12 octave smoothing - Amplitude Axis @ 6dB per division.

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.

Measurement Issues - Reflections in Impulse Responses.
Measurement Issues - Lies, Damned Lies & Statistics.
Acoustic Line Source Research - Table of Contents.