Figure 1. The same measurement presented with three different scales on the polar plot. Click here to see an enlarged diagram.
Standardization
A microphone is basically a simple device — a transducer that transforms vibrating air to an electrical signal. How come so many different microphone types and brands exist? Well, almost no two microphones are similar, primarily because microphones may be considered as special tools each selected for special jobs, and because of different design approaches among manufacturers. So reading specifications is important.
Some specifications are, of course, easily understood and even comparable between different brands. But, this may not necessarily apply to all specifications.
Sensitivity
International standards are written to provide a common basis for the specs. One of the most important standards is the “IEC 60268-4 Microphones,” a rather comprehensive standard. However, almost no manufacturer is following these guidelines from one end to the other. Hence, each manufacturer selects the specifications it wants to publish. Let's walk through a few of the most important specifications generally presented: sensitivity, impedance, self-noise, the maximum permissible peak SPL, frequency response and directional characteristics.
The sensitivity expresses the microphone's electric output when placed in a given sound field and a given sound pressure. The sound pressure is nominally 1 Pa (pascal is the unit for pressure). This is equivalent to a sound pressure level (SPL) of 94dB. If no further conditions are mentioned, the sensitivity is measured on-axis in a free sound field at 1kHz with (virtually) no load on the output terminals.
The sensitivity is expressed either in volts/Pa (in practice 1/1000 volt, mV) or in dBV/Pa. (���dBV” is the same as “dB relative to 1 volt”). This is an example:
10 mV/Pa, or -40dBV/Pa, or -40dB re 1 volt/Pa
If the microphone is placed in an SPL of 114dB, the output is 10 times higher (equivalent to +20 dB), which yields 100mV or -20 dBV.
If the microphones are individually calibrated, a calibration chart provides the exact sensitivity. If the data is taken from a catalogue, a ±1.5dB to ±2.5dB deviation from the nominal value can be expected. When selecting microphones for stereo pick-up, it is good practice to purchase a matched pair.
Impedance
It should be noted that sensitivity might be measured under other conditions, i.e. in the near field, in a diffuse field and at other frequencies than 1kHz. Also, the sensitivity may change (dramatically) when the microphone is loaded by the actual input stage (or input stages if the microphone feeds more than one input).
Self-noise
The impedance is defined as the internal impedance measured between output terminals. As it is common (at least in the USA) to let one microphone directly feed more inputs, the minimum-permitted load impedance can be stated. Note that a heavy load on the output of a microphone normally reduces its ability to handle high sound pressure levels.
All microphones have a noise floor. The basic noise is simply caused by the presence of air around the microphone due to the movement of the air molecules.
The noise specification is normally expressed as the so-called equivalent noise level. This indicates the sound pressure that will create the same voltage as the self-noise from the microphone produces. The rms-based measurement is A-weighted, meaning a filter basically subtracting less audible frequencies is used.
Figure 2. The frequency response on-axis (0°) and off-axis (30°, 60°, 90° and 180°). Click here to see an enlarged diagram.
Also, a CCIR filter (or ITU-R BS.468-4) is specified in connection with a quasi-peak level reading. This provides “worse” figures compared to the A-weight/rms, hence these data are not always presented. Here is an example:
Equivalent noise, rms, A-weighted: 15dB
or
quasi-peak, CCIR-weighted: 27dB
The maximum permissible peak SPL
In addition to a simple noise-figure, a noise spectrum also can be provided.
Frequency response
The IEC-standard defines this as “the maximum instantaneous sound pressure of a plane sound wave, specified by the manufacturer, that the microphone can tolerate without a permanent change of its performance characteristics, for any direction of sound incidence.” This is to some degree expressed differently by different manufacturers. So the conditions regarding these specs should always be read carefully. Is it the sound pressure that destroys the microphone, or is it just the sound pressure that causes a certain amount of distortion?
Directional characteristics
Often only the on-axis response is presented either as a curve (preferred) or as a frequency range within specified limits. However, off-axis responses can be of great interest, especially regarding directional microphones. Unfortunately, it is difficult to perform reliable off-axis measurements in the nulls of a microphone. So if it is not shown in the specs sheet, it is not necessarily bad will; it may have practical reasons. However, when reading the frequency response curves, be aware of the measurement distance (directional microphones) and the dB scale.
Check it out
The directional characteristics are expressed either by the characteristic pattern (omni, cardioid, figure eight, etc.) or by a set of curves. The curves are always interesting, and again the scale is important. (See Figures 1 and 2.)
More specifications could be mentioned. You can check out the manufacturers' Web sites or their catalogues. For instance, check out the Microphone University at the www.dpamicrophones.com, where more explanations can be found.
In order to make life easier to users, an important work currently takes place within the standards committee of the Audio Engineering Society (AES). This work addresses the problem concerning the non-comparable specification sheets. Ten of the leading microphone manufacturers each have provided one standard microphone, which is then tested by the nine others. This work is unique as measurements on competitors' products are reported and compared in order to provide the most correct and informative data to the user. The work is expected to be finished shortly.
Eddy B. Brixen is a consultant with EBB-consult in Denmark.
