Copyright © 1994-2002 by Rick Chinn. All rights reserved. Additional material © 2006 by Rick Chinn. All rights reserved.

Overview

MS-Stereo recording uses a figure-8 pattern microphone in conjunction with a cardioid pattern microphone, mounted as a coincident pair. The cardioid microphone (mid) faces straight ahead, towards the sound source. The figure-8 microphone (side) faces sideways, so that the null in the figure-8 pattern coincides with 0 degrees on the cardioid's polar plot. The resulting signals from the microphones correspond to the sum (cardioid) and difference (figure-8) of the left and right channels of a stereo pair. By summing and differencing these signals, the resultant signals correspond to the left and right channels of a pair of crossed cardioid microphones in the same location as the MS coincident pair.

If M = mono = L + R
and
S= difference = L - R
then by adding and subtracting the M and S signals
M + S = (L+R) + (L-R) = 2L
M - S = (L+R) - (L-R) = L + R - L + R = 2R

Varying the ratio of the side to the mid signal varies the width of the stereo stage. By recording the mid and signals separately, you can delay making the width decision until post production. Since the two signals are picked up with a coincident pair, and since the side signal cancels when the X/Y signal is mono-summed, an MS-stereo recording is inherently very mono compatible.

Although the two microphones are usually capacitor (condenser) types, there is no reason why the two microphones couldn't be ribbon or dynamic (or a mixture) units.

AKG, Neumann, Beyer, Shure, and Sony (there are probably others, too) all make stereo microphones (two microphone capsules mounted in the same body) that can be used for MS-stereo technique. Likewise, it is perfectly fine to use two discrete, separate, microphones as long as the two capsules are as coincident as possible.

A crude, but effective decoder scheme requires multing the side signal, and applying it to two inputs on the mixing console, panned oppositely, with the polarity switch on one of the channels engaged, so that the two channels are in reverse polarity to each other. The mid signal is then applied to a third channel, and this channel is panned center. You vary the stereo width by varying the level of the side signals simultaneously and equally on the two console channels that they are connected to.

Purists may choose to retch, as one of the microphones sees twice the load as the other, and the degree of matching between console channels may leave something to be desired. If your console's line inputs allow polarity switching, then you could refine this technique by multing after the mike preamps. The important thing is that you already most likely have the tools to experiment with this important technique without spending a dime. If this works for you, then you can give thought to trading dollars for gear.

Passive Matrixing

A passive circuit, using transformers, is presented in references 14 and 16. Here's three versions that are online: Neumann MS Stereo Matrix circuit  Jensen Transformers AS065  Jensen Transformers AS055 

The first link is the circuit included with the Neumann SM-2 stereo microphone. There are two circuits shown here, the first using a Wheatstone bridge to accomplish the sum/difference operation. Notice that the outputs are unbalanced. Also note the tolerance of the resistors, 0.1%. No the decimal point isn't misplaced, you need that level of accuracy to ensure that the two signals are accurately added and subtracted. Note that it isn't value accuracy you need, but matching between the four resistors. The 2nd Neuman circuit uses transformers to sum and difference the two signals. The circuit description for the Jensen circuits will suffice for this as well.

In both Jensen circuits, each microphone signal (M and S) is divided into two independent signals that are then summed (added) and differenced (subtracted). The transformers allow this because each signal is floating in space, and free to be connected in series aiding (adding) or series opposing (subtracting) with its opposite counterpart. The circuit that Neumann presented years ago (ref 20, 21) used two transformers that they sold for a princely sum (nearly as much as the microphone itself) and added a pair of T-attenuators at the inputs so that you could vary the matrixing. There is nothing wrong with doing the matrixing this way, although I consider using operational amplifiers to be a more elegant solution, and one that doesn't require nearly as much financial output.

Active Matrixing

The three circuits presented here solve the problem by using operational amplifiers to handle the summing and differencing. Each circuit is a bit more complex than its predecessor, and the third circuit is a pretty elegant solution to the problem.

Simple Circuit

Refer to the drawing: MS-MAT02.  This shows the basic intent of the circuit. The top opamp is a diff-amp that subtracts the L+R and L-R signals; the result is 2R. The bottom opamp is a summing amp that adds the L+R and L-R signals; the result is 2L.

This circuit works, but has a difficulty: the mid and side inputs should really be driven from a low source impedance. Typically you put pots between your MS-source and these inputs so that you can adjust the matrix coefficients (sum and difference levels) for the desired amount of stereo width. This makes the source impedance seen by these inputs non-zero. The effect is that the matrix does not strictly obey the model unless the model takes the source impedance into account. In practice, if the pots are 10K or less, then the worst-case source impedance seen by these inputs is 2.5 kilohms, which makes this (probably) academic.

The final difficulty is that the circuit phase inverts. Of course, for two more opamps or by simply inverting the polarity at the input XLR connectors...

More Complex Circuit

Now, refer to the drawing: MS-MAT01.  This shows a more refined version of the circuit. The two diff-amps have the same noise gain, and thus their noise-outputs should be more equal than the simple circuit. It does not invert phase. I feel it is a better circuit. If you put either circuit into the channel-inserts of a mixer, you can do away with the mid and side adjustment pots by simply using the gain-trim controls on the mixer.

A Deluxe MS-Decoder

Refer to the drawing in MS-MAT03.  This circuit combines the simple ms-matrix and a stereo-width control. The width control adjusts the L-R ratio of the stereo signal from mono through normal stereo in the middle to super-stereo at the other extreme. At least one commercially made ms-matrix has this feature. Some additional construction details are given as well.

General Notes for All Circuits

The opamps can be 5532's or other unity-gain stable opamp. You're on your own for bypassing, grounding, and any loop-rolloff capacitors. The circuit diagrams are also available as postscript dumps from my ftp site: 10/99: the ftp site is down while I change ISPs. if you want this, send email.

Another Twist

Lest you think that this technique is only applicable if you have an MS microphone, an ordinary LR stereo signal can be converted into its sum and difference components, operated on somehow, and then rematrixed back into LR stereo.

To get the sum (mid) signal, mix left and right together in equal amounts. To get the difference (side) signal, mix left and right together in equal amounts, but engage the polarity switch on the right channel. If you bug me about it, I could be presuaded to add the basic circuit to the library here.

Lest you think this is new, the much desired, usually overpriced Fairchild 660 limiter used this technique in the 1960s as a way to control the cutting stylus motion in a disk recording chain.

M-S Bibliography and Additional Reading

If the circuits don't keep you busy, then these references will. The references to/from the Journal of the Audio Engineering Society can be purchased online from the Society: www.aes.org 

  1. Basic Stereo Microphone Perspectives -- A Review, Ron Streicher and Wes Dooley, J. Audio Eng. Soc., 1985 July/August.
  2. M-S Stereo: A Powerful Technique for Working in Stereo. Wes Dooley and Ron Streicher, J. Audio Eng. Soc., 1982 October.
  3. Blumlein, British patent 394325, 1931 Dec 14, reprinted in J. Audio Eng. Soc., 1958 April.
  4. A Phased Array, HiFi News & Record Review, 1981 July.
  5. Recording in 2 and 4 channels, J. Gordon, Audio, 1973 December.
  6. Miking with the 3-Point System, Audio, 1975 December.
  7. Early Hi-Fi and Stereo Recording at Bell Laboratories, A. C. Keller, J. Audio Eng. Soc., 1981 April.
  8. Sound Recording, John Eargle, Van Nostrand Reinhold, New York, 1976.
  9. The Microphone Handbook, John Eargle, Elar Publishing, New York, 1982.
  10. Stereo Microphone Techniques: Are the Purists Wrong?, S. P. Lipshitz, Audio Engineering Society 78th Convention preprint #2261.
  11. Microphones-An Anthology, Audio Engineering Society, New York, 1979.
  12. The technique of the Sound Studio, Alec Nisbett, Hastings House, New York, 1974.
  13. Music, Physics and Engineering, 2nd Edition, Harry F. Olson, Dover, New York, 1967.
  14. Data Sheet for the JE-MB-D transformer, Jensen Transformers, Chatsworth, CA.
  15. Application note 065,  Jensen Transformers, Chatsworth, CA.
  16. Application note 055,  Jensen Transformers, Chatsworth, CA.
  17. Another Purist Technique, T. Faulkner, Hif-Fi News and Record Review, 1980 August.
  18. M-S Stereo Recording Techniques, B. Weingartner, North American Phillips Corporation (AKG distributor around 1960-1970).
  19. Application Notes, dB magazine, 1980 December.
  20. Principles of M-S Recording, Ron Streicher, Audio Engineering Associates, Pasadena, CA.
  21. M-S Stereophony and Compatibility, Gerhart Bore and Stephen F. Temmer, Audio, 1958 April.
  22. Neumann SM69 Double Condenser Microphone data sheet, Georg Neumann GMBH, Berlin. (see earlier link in website text).
  23. Neumann SM-2 Instruction Manual, Gotham Audio Corporation, New York. (see earlier link in website text)
  24. Neumann Information 90 013 
    Conversion MS-stereo signal to LR-stereo signal. German and English.
  25. An Intuitive View of Coincident Stereo Microphones, Stephen Julstrum, AES preprint 2984, 89th Convention, 1990 September.
  26. XY and MS Microphone Techniques in Comparison, M. Hibbing, J. Audio Eng. Soc., 1989 October.
  27. A Condenser Stereo Microphone with High Directivity Index, S. Peus, AES Preprint 2462, 65th Convention, 1987 March.
  28. Stereo Sound Recording with Shotgun Microphones, H. Gerlach, J. Audio Eng. Soc, 1090 October.
  29. New Perspectiveson Coincident and semi-Coincident Microphone arrays, D. Greisinger, AES preprint 2464, 82nd Convention, 1987 March.
  30. Recording Techniques for Multichannel Stereo, Michael Gerzon, British Kinematography Sound and Television, 1971 July.
  31. Stabilising Stereo Images, Michael Gerzon, Studio Sound, 1974 December.

Trademark notice: Some of the names used in this document are trademarks or service marks of various companies around the world. Their use here in no way alters their ownership of same. It is simply an acknowledgement of their existence, and the giving of credit where credit is due.

Last modified 1/8/2006.