More on Microphones by Michael Williams, (www.williamsmmad.com)   △   <   ∧   >

8.2 - DIRECTIVITY

8.2.6 - THE GRAPHICAL REPRESENTATION OF DIRECTIVITY

I suppose this section is really a request to the manufacturers to give us, the professional user, access to more useful information on the measured characteristics of microphones in general, and in particular the directivity response. We have become used to seeing the polar diagram response as the main graphical representation of directivity. Unfortunately in the specification sheets published by manufacturers, these diagrams are becoming smaller and smaller, and are being simplified to such an extent that they are becoming almost meaningless. The specific frequencies at which polar diagrams are measured have been reduced to the absolute minimum – the preferred frequencies being 125Hz, 250Hz, 500Hz, 1kHz, 2kHz, 4kHz, 8kHz and 16kHz. With any basic directivity pattern, the most significant divergence from the typical directivity response occurs at each extremity of the frequency band. Below 250Hz and above 8kHz, the standard polar diagram is in no way detailed enough to supply any significant information.

It is standard practice to measure the directivity response at these preferred single frequencies and present the result as a polar diagram. This can give a very distorted view of the overall response in different parts of the audible frequency range. It is not uncommon to see, in the cardioid and hypercardioid directivities, polar diagram traces that produce such tight null points that they could be confused with the theoretical model diagram as shown in Figure 49 in the previous section. In reality this is only valid for a few specific frequencies, and the response around these frequencies is far from being so neat. The scale used for showing the amplitude of response can also change considerably our visual impression of directivity – in Figure 51 the same microphone is shown with two different scales, 50db on the left and 25db on the right. The 50dB scale allows us to see more detail, but in fact the 25dB scale is a better match between our visual and our auditory impression. The 25dB scale is now almost universally used for polar diagram presentation.


Figure 51 – Supercardioid Microphone Polar Diagrams
with 50dB and 25dB Scales Measured at the Prefered
Single Frequency (except 16kHz)
© M160 polar diagrams published by courtesy of Beyer
For many years I have been organizing practical training sessions concerning microphones and microphone characteristics. In the early stages, students are required to draw by ear their appreciation of the various directivity patterns for specific types of microphones using a sound source that moves round the microphone. As the capacity of the student to discriminate between different parts of the frequency spectrum improves, it is possible to divide the spectrum into regions such as very low, low, medium, high and very high frequencies, and to draw the corresponding polar diagrams.

These results were then compared with the measured results, either as published or as measured in the laboratory. This exercise is highly instructive concerning the way in which we perceive the directivity response. From these results it would seem that octave band measurements of directivity as shown in Figure 52 (same as Figure 41 in section 8.2.1) correspond more closely to our perception of directivity, except for the higher frequencies from about 8kHz and above, where the 1/3 octave band measurements would seem more appropriate. I have found no examples of single frequency measurements, especially with tight null points, as being anywhere near the perceived results. The single frequency measurement is certainly of interest in the design and development of a microphone, but has little to recommend itself to the operational sound recording engineer.


Figure 52 – Octave Band Measurements of the Directivity Patterns
The standard amplitude/frequency plot as shown in Figure 53, made at different angles with respect to the microphone’s directivity axis, is sometimes more indicative of the character of directivity response in the low, medium and high frequency range, and can also point to some notable modifications in the basic directivity pattern at very low and at very high frequencies.


Figure 53 – Amplitude/Frequency Resqponse
for a Cardioid Condenser Microphone
The amplitude/frequency response is shown at 0°, 90° and 180° with respect to the microphone’s directivity axis. Please note that the dB scaling is from 0dB to 25dB and on double width paper which has been used to accentuate the detail in the frequency response curve.

The roughly parallel response from 200Hz to 5kHz between the 0° and 90° response curves, corresponds quite closely to the expected basic cardioid response – 6dB attenuation at 90°. Below 100Hz this microphone has a tendency to become bi- directional, whilst above 12kHz the response would be better described as hypocardioid. Between 5kHz and 12kHz the side attenuation of 6dB associated with a typical cardioid microphone is reduced to almost 2dB. The ‘chaotic’ response at 180° is typical of the pressure-gradient cardioid labyrinth design. In the general profile of this 180° response, we can see sharp null points corresponding to integer multiples of the wavelength with respect to the equivalent path delay introduced by the labyrinth structure, usually called a ‘comb filter’ frequency response.