FIG. 1: M-S miking combines a cardioid mic with a coincident figure-8 mic aimed to the left. A simple matrix combines the two signals to derive discrete left and right channels.
Photo: Chuck Dahmer
The mid-side (M-S) stereo-miking technique is well known in both classical and commercial recording for its ability to control stereo width while maintaining mono compatibility. I'll show you how to adapt M-S principles to bring that same ability, and more, to your mastering.
Adding It Up
M-S miking is a coincident technique combining a cardioid mic with a side-facing bidirectional mic (see Fig. 1). Because the bidirectional mic's null faces the sound source, it misses much of the direct sound and picks up sound primarily from the left and right sides of the stage. Thus, the cardioid mic is the mid, and the bidirectional mic is the side. Increase the level of the mid, and the stereo image narrows; increase the level of the side, and the image widens. Because the two mics are coincident, they sum to mono very nicely.
You derive discrete left and right stereo channels by adding and subtracting the two mic signals. That works because when we face the side mic to the left, sound from the right of the stage is picked up by the rear of the mic, pushing the diaphragm when it should pull, and vice versa. Therefore, the right-side signal has its polarity inverted — a negative of the actual sound. That means the side-mic signal is L - R, whereas the mid-mic signal is L + R. Sum those, and you isolate the left side; subtract them, and you isolate the right side.
To manipulate stereo width in mixing or mastering, you need to reverse the process and derive mid and side from left and right. Sum the left and right channels in your mixer, and you have the mid channel (L + R); sum them after flipping the polarity of the right channel, and you have the side channel (L - R).
To do that, copy the left and right signals, and invert the polarity (often incorrectly called the “phase”) of one copy of the right channel. This is easily done on many analog mixers, but DAWs often lack phase-invert buttons, in which case you'll need to use a plug-in such as the free Sonalksis FreeG (sonalksis.com). There are dedicated hardware and software processors that offer integrated M-S processing. To create an M-S matrix in Digidesign Pro Tools, see “Step-by-Step Instructions.”
Because you're creating parallel signal paths, delay compensation may be required, depending on your DAW. In Pro Tools LE and M-Powered, there is no intelligent automatic delay compensation, but buffering at the CPU keeps things aligned perfectly under many circumstances. If you use a plug-in with its own look-ahead buffer on the mid or side track, however, you'll need to use Time Adjuster to delay the other track by the same number of samples. In DAWs that offer it, such as Pro Tools HD, simply enable automatic delay compensation.
Note that in the example project (see Web Clip 1), the mid and side tracks are trimmed down by 6 dB to prevent clipping when the left and right channels are summed to a mono bus. Feel free to hide the left and right aux tracks in the example when you're done, as you won't need to manipulate them further.
Squeeze from the Middle
Once you've derived mid and side, controlling width is only half the fun. Create master faders for the buses carrying the mid (bus 5) and side (bus 6), and use their inserts to process them separately. A favorite mastering trick is to use the mid channel to EQ the lead vocal; because it's usually panned to center, you won't affect the rest of the mix. You can also use your favorite multiband compressor to rein in the bass and kick on the mid channel without affecting the higher-frequency lead vocal in the mid channel and without altering any of the low sounds in the side channel.
If you want, you can go back and ride the fader on the lead vocal. To do that, split the mid channel to a pair of mono aux tracks, then use complementary highpass and lowpass filters on the two tracks to isolate the vocal on one of them.
Compress the side channel, which usually contains most of the ambience in a track, to correct a mix that's too dry without adding another layer of reverb. Because the bass and kick are usually panned to center, rolling off some of the low end of the side channel helps clear space for those two critical components.
Use a tube-emulation plug-in to warm up the mid channel without clouding the rest of the mix. Sonic dissection opens up all sorts of imaginative possibilities. For example, try the technique on stereo drum overheads, room mics, or any other stereo track or submix. Once you've mastered the mid-side matrix, anything is possible.
Brian Smithers, author of Mixing in Pro Tools: Skill Pack (Cengage Learning, 2006), teaches at Full Sail University, where he is chair of the Workstations Department.
Step-by-step instructions on page 2
Step 1: Return the stereo mix on Bus 1-2 to the two stereo aux inputs called Mid and Side.
Step 2: Use multimono Trim plug-ins to trim both auxes by 6 dB and to invert the right channel of the Side track. You'll need to unlink the left and right channels of the plug-in to do this.
Step 3: Assign the outputs of Mid and Side to Bus 3 and Bus 4 to accomplish the summing (L + R) and subtracting (L - R).
Step 4: Return Mid and Side (Bus 3-4) to the two stereo auxes called Left and Right.
Step 5: Use a multimono Trim plug-in to invert the right channel of the right aux input. Assign their outputs to Bus 5 and Bus 6 to accomplish the summing (M + S) and subtracting (M - S).
Step 6: Return Left and Right (Bus 5-6) to a stereo aux. Adjust the level of the Mid aux to vary the stereo width.