A Series Active Monitor Speakers
E4/36 Service Date printed: 29.02.00
4.6 Appendix: Frequency Response Measurements
(Most parts of the following paragraphs are reprinted with kind permission
of Audio Precision, Inc., http://www.audioprecision.com)
4.6.1 General
Frequency response measurement of loudspeakers is a much more complex
topic than response measurements on amplifiers or most other electronic
audio products. Nearfield effects cause the response measured with close
microphone spacing to vary from that with distant microphone placement.
Loudspeakers are directional, so the response varies as the microphone is
moved off the speaker axis. Unless the loudspeaker is being measured in an
expensive anechoic chamber or in essentially free space outdoors, reflections
of the sound waves will also arrive at the microphone and create standing
wave patterns with cancellations and reinforcements. Multi-way loudspeaker
systems, ported loudspeaker systems, and systems with some of the drivers
facing the sides or rear of the cabinet all create a problem for measurement
microphone placement – just where should the microphone be located so
that it receives acoustical output which is a proper combination of the vari-
ous drivers and ports? Furthermore, the end use of the speaker will not be in
an anechoic chamber or laying on its back outdoors with listeners suspended
above it – so what is the proper environment in which to measure it?
4.6.2 Standing Waves, Reflections, and Smoothing
Neither an anechoic chamber nor outdoor free-space testing are practical for
production testing. In most test locations, acoustical reflections and ambient
acoustical noise are facts of life. Particularly at the higher frequencies, the
acoustical energy arriving at the microphone is a combination of the direct
signal from speaker to microphone plus a large number of reflections from
walls, ceiling, floor, human operators, etc. In a typical office or laboratory
environment when using high-frequency sinewave signals (10 kHz, for ex-
ample), movement of a human being several feet away from the microphone
can cause several dB change in the measured value. Changes in the order of
a few inches in microphone or loudspeaker placement can cause response
curve variations of many dB at high frequencies. Even at fixed speaker and
microphone locations and with no people moving within a reasonable dis-
tance, a response curve may show individual sharp peaks and dips of
10…20 dB due to standing wave effects.
For reasonable results, some degree of acoustical treatment of the measure-
ment location should be done. Commercially-available sound absorption
materials should be applied to wall, ceiling, and floor surfaces. It may be ef-
fective to design the shape of the test chamber as something other than a
rectangle to avoid high-intensity single-bounce reflections from the back
wall into the measurement microphone.
Since it is not economically practical to provide a true anechoic chamber for
most testing, speaker measurements in a reflective space are almost invaria-
To Next Page
Loading...
Need help?
General