The Mid-Sides Myth

While M‑S miking does offer you some flexibility at mixdown in terms of stereo width, it can’t perform miracles!

The most common way of setting up two microphones to capture stereo involves using one mic for the left channel and the other for the right channel, a scenario often called L‑R stereo. As many readers will know, though, there’s another way of capturing stereo with only two mics: the Mid‑Sides format, where one forward‑facing ‘Middle’ microphone captures the centre of the stereo picture, and a second sideways‑oriented figure‑8 ‘Sides’ microphone provides the stereo spread. The technical details of how M‑S stereo miking works, as well as how you can convert back and forth losslessly between M‑S and L‑R signal formats, were explained in great detail by Hugh Robjohns in SOS October 2025’s instalment of his excellent This Is Stereo series (now also available as a free SOS eBook: https://sosm.ag/stereo-miking-guide), so I won’t go back over all of that here.

What I would like to do, though, is tackle what I think is a common misconception amongst project‑studio operators (and perhaps even some more seasoned engineers) regarding the M‑S mic technique: namely, that you don’t have to worry about the Stereo Recording Angle (SRA, ie. the physical angle of capture in the recording room) of an M‑S array while tracking, because you can freely adjust the Middle and Sides levels of the stereo recording to set the desired image width at mixdown.

In reality, M‑S rebalancing of any stereo recording is fraught with compromise, and I’ve lost count of the number of times I’ve heard stereo recordings that have been impaired simply because the engineer treated M‑S as some kind of ‘Get Out Of Jail Free’ card! Below, with the help of some audio examples that I created specially for the purpose (https://sosm.ag/the-ms-myth), I’m going to demonstrate some of the important limitations of M‑S rebalancing.

The Myth

The first thing to realise is that you can freely convert any stereo file between M‑S and L‑R formats using a simple M‑S encoder/decoder without any change to the eventual mixed result. So, as far as your stereo mixdown is concerned, raising the level of the Sides mic in an M‑S stereo mic rig will cause exactly the same kinds of audible changes as encoding an X‑Y stereo mic rig into M‑S format and increasing the Sides signal level there.

With this in mind, have a listen to my first audio example (msm_01_FullWidthPickup), which was recorded with a crossed pair of Shure KSM 141 small‑diaphragm cardioid capacitor microphones, set up at a mutual angle (the angle between the two mics) of 110 degrees. The sound source for this recording was a stand‑mounted single‑driver loudspeaker that I physically moved to different positions around the mic rig, and which played an identical short spoken phrase in each position.

On account of careful positioning of the loudspeaker in the recording room, keeping it equidistant from the microphones at all times, this audio file should present you with the following stereo experience if you’re auditioning from the sweet spot of a stereo loudspeaker system:

• Phrase 1 will be at the centre of the stereo image.
• Phrase 2 will be at the far left of the stereo image.
• Phrase 3 will be at the centre of the stereo image again.
• Phrase 4 will be at the far right of the stereo image.
• Phrase 5 will be at the centre of the stereo image again.

Furthermore, notice that the perceived level of the spoken phrase remains pretty much the same as it moves around the stereo image, and that it also maintains roughly the same subjective distance from the listener.

Next, check out my second audio example (msm_02_NarrowPickup), which was made with the identical recording setup, but with the sound‑source loudspeaker positions adjusted so as to deliver much less extreme stereo positions for Phrase 2 and Phrase 4, while still maintaining a fairly consistent relative balance and perceived distance for all five phrases. I’d like to stress that these changes were created solely by moving the source loudspeaker in the recording room. I didn’t change the microphone setup at all. I didn’t adjust anything in the mix. Both audio examples comprise just the raw signals from the microphones, panned hard left and hard right, respectively.

Voxengo’s freeware MSED plug‑in is great for simple Mid‑Sides rebalancing of stereo recordings — just so long as you’re aware that it won’t just be adjusting stereo width, but also ensemble balance, depth, and reverberance.

M‑S Widening

Now, listen to my third audio example (msm_03_NarrowPickupWidened), in which I used simple M‑S rebalancing to widen my second audio example, so that the image positions of its spoken phrases match those in my first audio example — to put it another way, I tried to adapt the mic rig’s Stereo Recording Angle after the fact, to match the narrowed positions of the sound‑source loudspeaker in the recording room. You’ll hear that, yes, I have managed to ‘correct’ the stereo image as planned. Crucially, though, that’s not all that the process changed.

Firstly, you’ll notice that the balance between the left/right sources and the central source has substantially altered: the left/right sources in the...