MS Stereo

Maybe Phillipe can offer some enlightenment here. I record MS into Cubase by using the normal two mic (1 off axis) to a stereo track to give me a mid/side track I can edit, this is always guitar. I then take the finished edited stereo file and split it in WL, make an inverted version of the side and then make a new stereo file by adding the inverted side + mid for the right channel, and the original side + mid for the left channel. The result is always very good at capturing the recording space but a bit long winded.

I assume tools one does this automatically from the original file saving all the messing around, I have a couple of RME Firefaces that can do M/S as well. What I don’t get is how you know what the plugin is doing, could it be inverting the mid and adding that to the side instead of the other way round. Come to think of it would that make any difference apart from maybe (or maybe not) reversing the stereo field? So in simple terms what is the plugin actually doing?

M-S as you appear to understand represents Mid(dle) or maybe “mono”, and Side (the way the fig-8 mic points). M-S can also be created from an normal L and R stereo signal, by doing what you normally do to get a mono signal, that is, add (or sum) the L and R channels together. If you subtract the R from the L signal, you will get the difference between the two, that is the “stereo component” or what makes the L channel different from the right to create the stereo effect. The usual M-S mic arrangement of cardioid (for mono mic) and fig-8 for side or difference mic does this naturally. (Remember that in the fig-8 mic, sound from the right will have the opposite polarity to that from the left).

Now in converting an M-S signal back to L/R, all you have to do is exactly the same process. If you add the M and S signal in equal proportion, the L difference components in the S signal re-inforce the L component in the mono signal,but will tend to cancel the R components. Subtracting the S from the M will do exactly the opposite to produce the R signal again. This is precisely what you do in your mixing configuration, sometimes known as the “three-fader trick” which is rather cumbersome as you say.

What you may also want to do in processing an M-S signal may be to change the apparent width of the decoded signal. The more side signal you add, the more you will accentuate the differences between L and R, producing a “wider” and more “spacious” image (but don’t go too far!); less side signal will decrease the differences and you will end up with a narrower stereo image. Of course once you have decoded the signal to L and R you may also want to balance or pan the L and R signal to position the image to your requirements.

All the above is what an M-S plug-in can do. Others occasionally offer the ability to apply different processing to the “side” chain for special effects, but the main function is the “sum-and-difference” processing described above. There are many M-S plug-ins available. My current preference is the Voxengo MSED (M-S EDitor) which is a free download (with excellent documentation). No more cumbersome processing as you describe.

M-S is a great technique with many applications. Enjoy.

Thanks for the detailed reply. So by using the MS plug on my two channel “stereo” file containing my mid and side I will get a stereo file the same as using the 3 fader method. And I assume that being the case I can vary the stereo width by simply changing the pan position of the ms “stereo” file. A lot easier than the method I use now!

Hi, Folks!

What many folks don’t realize is that a Cardioid microphone pickup pattern can actually be created by the sum (or difference) from the output signals of a closely spaced or coincident pair of microphones: an omnidirectional and figure-8.

What’s even more intriguing is a number of these techniques were pointed out in Alan Blumlein’s seminal 1931 British patent 394,325 on Improvements to Sound Systems.

Further, by specifically pointing the omni microphone at the sound source, and then coincidently placing the figure-8 microphone such that the “8” pattern is normal or at 90-degrees to the sound source (the top of the “8” points 90-degrees to the left, the bottom of the “8” points 90-degrees to the right of the sound source); that summing these two patterns provides a left facing cardioid, taking the electronic difference between these two patterns provides an inverted phase output for a right facing cardioid. Correcting for the phase inversion of one of the mixed pairs, automatically creates a back to back cardioid stereo pickup pattern, if the two signals are equivalent in output level. For more info. on this topic, see the late John Eargle’s excellent work: The Microphone Book.

As noted above, varying the level of either the Omni or Figure-8 microphone changes the “stereo width”, as the omni pattern alone becomes the “center” monophonic microphone, and the Figure-8 mixes provide the necessary information, when added or subtracted electronically from the omni pickup, to create a stereo image of varying “width” depending on how much of this microphone’s signals are mixed with the omni. Other pattern microphones may be substituted for the front facing omnidirectional, but this is the basic form for this Mid-Side, or M-S technique. Additionally, summing the two “back to back” extracted cardioid patterns together, simply creates an omnidirectional pattern with higher output level than the originals.

A Mid-Side decoder simply accomplishes all of this phase inversion and sum-difference “jiggery-pokery” with a couple of simple controls for the 2 microphone channels. A similar approach has been implemented by Schoeps with their KFM-360 system, invented by Jerry Bruck in 1997, to create a powerful, small, 4 channel microphone array for surround-sound pickup that is quite impressive! Check out the Schoeps website for further information on this technique:

In both instances of these microphone techniques, recording the mics straight through to a multi-channel recorder during a session, permits extensive post-production balance and mixing flexibility, well beyond what is normally achieved with a spaced pair, two or more microphone recordings. The difficult part for live recording is achieving an acceptable distance from the performers so one performer isn’t out of balance with another, and the overall balance for the performers and ambient acoustic is acceptable; as even with the flexibility of post-production mixes, improper mic placement too close or too far away from the performers can severely color the sound, creating too “close” an acoustic, or a “wallowy” reverberant one on playback.

Hope this proves helpful, and if you really dig into it, the above references, including the astonishing Blumlein patent (Which explains in detail how to both record and extract 2 channels of audio information from the media of the time: a black, 78 rpm disk; pre-dating the actual implementation of black-disk stereo records by about 25 years!) are very informative!

You must have been reading a recent post of mine in another forum. The microphone in question is a Western Electric 639A and its numerous international variants. Mind you, check out the wildly varying responses of the ribbon and omni elements then figure out what the cardioid sounded like.

FM stereo radio uses M-S encoding - the sum signal is carried as a normal mono signal (for compatibility with old mono FM sets) and the difference signal is transmitted as a subcarrier at a higher frequency which stereo receivers can recover and do sum and difference on the mono signal to recreate the stereo signal.

M-S transmission was also used on program lines in the early days of stereo transmission, when only mono program line were available, and each line of the stereo pair could be routed via a different path, leading to time as well as frequency differences. On transmission these differences could lead to strange effects when received on a mono receiver. Transmitting over such program lines in M-S meant that mono signal integrity could always be guaranteed. Eventually, stereo program transmission technology was developed which eliminated this issue.

Oh, it was tough in the bad old days …

Hi, Folks!

Mono-Stereo compatability is still an important issue, although less predominant than it used to be. Spaced omni stereo pickup on occasion can result in “comb” filtering at low frequencies, which can cause loss of bass when played back in mono (which can result on car radios when station signal strength drops, forcing monophonic reception). Hence, coincident microphone stereo and M-S techniques tend to preserve monophonic compatability, even though the “stereo width” can be reduced (however, using various additional microphone and M-S techniques can be used to enhance and/or restore more pleasing stereo “width” to the playback).

Even some of the early “quad” playback systems used such summing and subtracting techniques, similar to FM Stereo broadcasting, to create the surround channels, as did the Hafler Dyna-Quad system, which provides very impressive surround out of a crossed pair of coincident cardioid micorphones in the so-called X-Y technique.

Looks like Phillipe has given his answer in the form of a plug in, stereo tools, nice one.