The “Mid/Side”, or M-S recording technique, was first designed by EMI engineer Alan Blumlein. He invented the method in 1933 and used it for some of the first stereophonic recordings. The M-S microphone recording technique is commonly used in broadcasts, mainly because when properly recorded, M-S tracks are mono-compatible. M-S is also a common studio recording technique, and its simplicity and versatility even make it a good option for live recording.
The M-S recording technique consists of two microphones. One directional cardioid, or hyper-cardioid mic, M, and a bidirectional microphone, S. The former points to the source and the latter has its front and back pointing towards the sides, perpendicular to the directional mic’s angle. Microphones should be as close as possible without touching.
M-S has outstanding mono performance, and in the right conditions, it can be the best way to capture the space and ambient of a room while keeping compatibility. This can even sound more realistic than a separated pair in many situations. However, when the source is particularly big, it can often take too much of a distance to get the whole section or choir into perspective using this technique. Several spot mics should be used in such a case.
As general advice, for best placement, we should walk around the room and listen where the instrument or source sounds best. It’s important to note the balance of instrument and room sound and the stereo image as well. Once the right spot is found, we can set up the mics. The M should go first since it will be our main mic. The bidirectional comes after, to create the stereo image.
Decoding the M/S signal
A stereo speaker system needs a “decoding” method to correctly translate the M-S recording technique. It’s a consequence of how the bidirectional mic captures sound. The directional, on the one hand, produces a “positive” voltage all the way around. The bidirectional, however, generates a positive voltage from sources coming from its front and a negative voltage from its back, albeit the signal is still mono. If we intend to use it to add stereo wideness, we need to decode the signal in the following way:
On one channel of the mixer, we should bring up the “M” signal. Pan it dead center. Next, we need to bring up the “S” bidirectional on two additional channels. Exact copies, but panned hard left and hard right. One of these channels should have its phase flipped. Only then will it serve the purpose of adding stereo differences, thus creating stereo wideness. Both S channels can be then balanced as a group against the M signal for more or less effect.
Additionally devices such as “M-S encoders” exist, which allow for avoiding this manual decoding process.