We had used various commercial microphones and found omnidirectional and continuously variable D types to have satisfactory sound characteristics (naturalness) for the vocals. But since it was our desire to exclude the instrumental sound from the vocal system we had a problem which could not be solved by a directional microphone because the instrument loudspeakers are located behind the vocalists on axis with the desired pickup. Therefore we turned to the dipole type in which the response is a function of the difference of the sound pressures at two distinct points. This configuration today (1975) is usually called a differential microphone, and as a close-talking microphone, the output is independent of frequency. This is a first-order gradient microphone and it possesses excellent antinoise characteristics. We place the pressure microphones about 60 mm apart. Wider separations reduce the effectiveness for higher frequencies and closer spacings can roll-off the low frequency in voice since the low frequencies of the vocalist can be heard by both microphones.
Our first implementation was with two dynamic microphones connected in series opposition. When the input impedance is much higher than the geneator impedance of the microphone, each can generate its voltage properly and excellent results are obtained.
(We do not recommend the reverse-polarity parallel connection as the inductance of the second microphone will attenuate the low frequencies of the vocal microphone in the same manner as a "voice" response inductor which is switchable in many microphones to roll off the bass response.) Often, splitter connections must be made to the microphones for recording and live broadcasting and under these conditions, the simple connection shown which offers a relative independence of loading effects. By this time we had achieved pretty good rejection of our instrument sound field but we felt that with closer matched microphones we could do even better.
We tried condenser measuring microphones from various manufacturers but found only one which would give us better matching than the dynamic microphones. For the final system, we asked this manufacturer to select a series of elements matched in amplitude and phase and obtained elements with amplitude match better than ± 0.1db and phase match better than 1 degree at 10KHz! This precise match made our noise canceling performance outstanding and the sound quality of the microphone for music is unexcelled. We used preamplifiers designed to use the measuring microphone elements with an instrumentation tape recorder but modified them for greater dynamic range.
All the direct (vocal) microphones were resistively summed and all the ambiance (noise canceling) microphones were resistively summed and then the difference was taken by an ultra low distortion amplifier. No conventional gain controls were used and thus the signal path was kept as clean as possible. The sensitivity of the microphones is controlled by varying the polarizing voltage applied to the condenser element. The same voltage is applied to both microphones of the differential pair, preserving the noise-cancelling capability.
A control is mounted on as small box which serves as the mounting structure for the microphone. This control permits the performer to control the output of his microphone. Provisions are also made to remotely control the sensitivity when announcements are being made by persons unfamiliar with the system. On the remote control panel is a switch for each microphone which convert it to an omnidirectional microphone by bypassing the ambiance element.
For recording and broadcasting, each microphone is provided with separate difference amplifiers which has two transformer isolated outputs. Each microphone may then be recorded on a separate track.
Because of the antinoise characteristics of the microphones, the sensitivity drops off rapidly when one moves away from the element. This causes some problems for first-time users who must not wander around as they may be accustomed to with conventional microphones. There are those who reject the microphones for this reason. Heaters should be provided for the microphones since considerable condensation from the breath can load the diaphragm and destroy the match of the elements.
"If you get confused,...
Listen to the music play"