Middle-Side (MS) by z8OBCl


									                          Middle-Side (MS)

          Recording and Playback Techniques
                                      Ian Corbett

                      Sound Engineer and Music Technologist


            •MS Microphone Technique

            •Matrixing MS to Stereo

            •Why Use MS Techniques?

            •MS and the Broadcast Industry

            •Stereo Loudspeaker Arrays and Phantom Images

            •Matrixing Left-Right Stereo to MS

            •Playback Using a Bi-Polar Loudspeaker Array

            •Using Bi-Polar Playback Techniques in the Theater

            •Double MS

            •Suggested Microphones for MS Recording

MS Microphone Technique

            The diagram below shows the typical microphone set up for recording
            using the MS technique. A cardioid (condenser) microphone is used to
            pick up the Middle (mono, or front-back) information, and a figure-8
            (condenser) microphone is used to pick up the Side (left-right) information.
            This microphone array may be either two microphones placed as
            coincidentally as possible (to avoid phase discrepancies between the two
            diaphragms), or a dedicated dual capsule microphone.
              The M and S signals need to be ‘matrixed’ or decoded before one can
              listen to a stereo (left-right) signal. The matrixing is necessary to derive
              discrete left and right information from the S microphone, which is a single
              capsule — only one diaphragm. With the S microphone aimed + and -
              (referring to the front and rear of the capsule) as in the diagram above,
              the L information is produced by taking the sum of M and S, and the R
              information by producing the difference between M and S;

                                  L = M + S+ (or M + S)

                                   R = M + S- (or M - S)

              where S+ is the regular output of the figure-8 microphone, and S- is the
              output of the figure-8 microphone, but phase inverted by 180o.

Matrixing MS to Stereo

              Matrixing can either be done ‘manually’ on a mixing board, or by using a
              dedicated matrix decoder into which the M and S components are fed,
              and a stereo L-R signal is output (the relative levels of the M and S
              components being controlled by knobs on the decoder).

              Three mixer channels are required to matrix the signals manually. Into
              channel 1, feed the output of the M microphone, pan it center and label
              it ‘M’. Into channel 2 feed the output of the S microphone, pan in hard L
              and label it ‘S+’. Channel 3 is fed from the direct out of channel 2 (and
              therefore is at line level, post fader) and is phase reversed, either by
              engaging the phase reverse button, or using a phase reversed cable
              between channel 2’s direct out and channel 3’s line input. Label channel
              3 ‘S-’.

              For the MS matrixing to work properly it is essential that both the S+ and S-
              channels are at the same level. To balance them:

                             •mute the M channel (1),

                             •pan channel 3 hard left (the same as channel 2),

                             •adjust channel 2’s trim to a healthy level (about 6db less
                             signal than the M channel should be fine) and push the
                             fader up high (so you can really hear what is going on),
                             •set channel 3’s fader at unity. The same signal (but in
                             phase in channel 2, and phase reversed in channel 3) is
                             going to the same monitor speaker. Theoretically, when
                             both + and - are at the same level, complete cancellation
                             will occur, and no sound will be heard. So;

                             •adjust the trim level of channel 3 until the relative levels of
                             both channels are the same, and the sound is attenuated
                             as much as possible,

                             •pan channel 3 hard right. A ‘phasey’ confused sound
                             should result,

                             •reduce the increased level of channel 2, and un-mute
                             and bring in channel 1 (the M component) until a good
                             stereo image is produced.

              The S component can be adjusted in level by moving only the channel 2
              fader. (Channel 3 is being fed post pre-amp, and post fader so will also be
              brought up and down in level, maintaining the essential relative balance
              of the two S channels.) An accepted ratio for the signal strengths of the
              M:S components is 3:1.

Why Use MS?

              When using a typical X-Y coincident pair microphone array, the left and
              right information is coming from two distinct sources — each microphone,
              one aimed left, the other aimed right. These signals are then panned left
              and right to produce a stereo signal. If set up and engineered correctly
              this X-Y coincident pair technique can produce excellent results, however
              there are two problems associated with it. The first is that with the
              microphones set at 90o to each other, a lot of the same signal can be
              going to both of them, producing a very mono sounding recording. This
              can be remedied by increasing the angle of separation, however care
              must be taken not to create a 'hole' in the center of the stereo image as
              this angle is increased. The pick up pattern of the microphones in use will
              also dictate what angle of separation is needed to obtain optimum
              results. The second problem is the occurrence of phase problems due to
              the distance between the two capsules, particularly when not using a
              dual capsule microphone. X-Y spaced pair microphone techniques are
              prone to even more phase problems, and stereo imaging is generally
              more diffused.

              Little can be done to correct problems such as the width of the stereo
              image once an X-Y recording is committed to two-track tape (without
              using MS techniques described later in this document). MS signals can
              either be matrixed to L-R stereo information and recorded onto two track
              tape, or without matrixing, the M and S components can be recorded
              onto tape, and the matrixing done at a later date, maybe in better
              mixing/matrixing facility.

MS and the Broadcast Industry

              If a matrixed MS recorded signal is collapsed into mono, what happens?
              The S+ is summed with the S-. A 'plus' and 'minus' cancel out completely,
              so all that is left is the M component which carries all the important sound
              information, just not the ‘frosting’ supplied by the S component. An
              acceptable mono signal is produced, with no phase anomalies audibly
              present. The same cannot be said of collapsing an X-Y recorded signal
              into mono, a process which is prone to phase problems.

              AM radio is a mono broadcast medium, so the original recording engineer
              has to have produced a mono compatible mix. FM radio broadcasts L
              and R signals, and when listening on a mono radio the mono signal is
              produced by summing the L and R — since people still listen on mono
              radios the issue of mono compatibility is still a concern. Television, now a
              stereo medium, broadcasts in MS, the decoding occurring inside the
              stereo TV. Before the days of stereo TV, sound was just a mono (or 'M')
              signal. By adding an S component to this existing mono signal, stereo
              images are produced (matrixed) by suitably equipped TV sets — mono
              sets simply ignore the new S component information. Since most viewers
              have mono TV sets, mono compatability is (currently) of great concern to
              TV broadcasters.

Stereo Loudspeaker Arrays and Phantom Images

              The problem with listening using a conventional two loudspeaker stereo
              array is that most of the 'bread and butter' information is panned
              centrally. This means that there are equal amounts of the same signal
              going to both the left and right speakers, producing a 'phantom' image in
              the middle of the stereo sound-stage. The problem with this phantom
              image is that one has to be situated in the correct location to hear and
              place the image correctly in the middle of the speakers.
               An experiment to try: mic up a piano (mono), or take a mono recording of
               a piano, and listen to it panned center (listening from the 'sweet-spot') —
               this is a 'phantom' image, positioned in the space between the two
               loudspeakers. Next, pan the signal hard left or right, and listen to it from
               directly in front of the speaker. The single speaker, non-phantom image
               should sound crisper, and less fluffy. The reason for this is that when signals
               are traveling from the left and right speakers, sound from the left speaker
               is going to one's left ear, and sound from the right speaker is going to
               one's right ear. This is fine! What also happens is that sound from the left
               speaker goes to one's right ear, and sound from the right speaker goes to
               one's left ear, slightly time delayed due to the extra distance it has to
               travel around one's head. This causes phase cancellation and anomalies
               which distort the stereo image.

Matrixing Left-Right Stereo to MS

               It is possible to take any Left-Right stereo signal, and matrix it into it's M and
               S components. Why do this? Mastering engineers have used this
               technique to widen (or narrow) the stereo image of recordings that are
               too 'mono'. A stereo L-R signal can be narrowed by panning the L and R
               channels more centrally, but may cause phase cancellation artifacts if
               the signal is not fully mono compatible. It is impossible to widen the stereo
               image by using pan pots. Matrixing a L-R stereo signal into it's M and S
               components allows one to widen the stereo image by increasing the S
               component in relation to the M component, and also to listen to
               playbacks using a technique that eliminates the center phantom image
               problems described previously. One way of matrixing a L-R stereo signal to
               MS requires a mixing board with a mono output, an augmented speaker
               array (and a soldering iron to make the necessary cables):

                               •bring the Left and Right channels of the source material in
                               on two mixer channels,

                               •pan them hard left and right respectively, and assign
                               them to a pair of output busses (for example, Group outs 1
                               and 2),
                             •derive the M component by assigning both of these
                             channels to the mono output bus on the board.

              The S component requires a special cable to go from the two Group outs
              (as above) to a single BALANCED line input:

                             •connect the + wire of the LEFT/GROUP output (as
                             assigned above) to the TIP (or the + XLR pin) of the
                             balanced line input connector.

                             •connect the + wire of the RIGHT/GROUP output (as
                             assigned above) to the RING (or the - XLR pin) of the
                             balanced line input connector.

                             •connect both of the shields to the sleeve (or XLR pin 1) of
                             the input connector.

              This new plug is the S component.

              These M and S signals can either be brought into other channels of the
              same mixer and fed to the L-R outputs, matrixed to form L-R again in
              exactly the same way as described in the 'Matrixing MS to Stereo' portion
              of this document. The stereo width of the originally L-R source material can
              now be changed by varying the amount of S component relative to the
              M component.

              If the board you are using does not have a MONO output buss, you can
              assign the L and R inputs to two odd (or even) numbered Group Outs, and
              use a Y cable to sum these outputs into one input connector.

Playback Using a Bi-Polar Loudspeaker Array

              Using a bi-polar loudspeaker array to listen to MS material requires three
              loudspeakers, arranged as in figure 4 — center, left and right, but with the
              left and right speakers to either side of the listener's head. It is essential
              that the distance between the listener and each speaker is the same
              (making this technique somewhat unsuitable for the 'real world', but still a
              unique listening experience).

              To apply this technique to a non-matrixed MS recording:

                              •connect the center loudspeaker to the M component

                              •split the S component source into two lines,

                              •feed the L speaker with the non-phase reversed signal

                              •feed the R speaker with the phase reversed signal (S-).

              If one has matrixed a L-R stereo recording into it's M and S components
              (as in the section 'Matrixing MS to Stereo', above):

                              •assign each of the three input channels to a different
                              Group output buss (or assign the M to a Group out, and the
                              S+ and S- to the L-R outs respectively),

                              •feed the center speaker from the Group output buss that
                              the first input channel (M) is assigned to,

                              •feed the left speaker from the Group output buss that the
                              second input channel (S+) is assigned to,

                              •feed the right speaker from the Group output buss that
                              the third input channel (S-) is assigned to.

              Using this technique for listening, one has full control over the relative M:S
              levels, and experiences a sonic image that has more clarity and integrity
              than the equivalent L-R stereo image.

              The distance between the listener's head and the speakers will affect the
              size of the 'sweet-spot' proportionally. However this 'sweet-spot' is small, so
              there is little 'real world' application for this speaker technique.

Using Bi-Polar Playback Techniques in the Theater

              A large sonic space can be created by using bi-polar speaker arrays in
              the theater. Distinct L-R panning will not be achievable due to the few
              number of people sitting in a spot equidistant to all speakers, however
              ambiances work very well. This technique is particularly useful if the
              production is 'in the round' or with seats on either side of a long central
              stage. It can also be applied to more conventional spaces as well.
             The M component is sent to a speaker array above the central stage
             area. S+ and S- speakers are placed in each corner of the room,
             alternating S+, S- around the room.

             In a more conventional space, M speakers could be hung center stage,
             and S+ speakers in one rear corner of the hall, S- speakers in the other.

Double MS

             Double MS microphone technique was developed during the 1980s. It
             involves using two MS microphone arrays, one for direct sound, and a
             second array to pick up room ambience and reverberation — usually
             positioned facing the rear of the room at or just beyond the critical
             distance of the room. (Critical distance being the point at which the
             direct sound and reflected sound are at approximately equal intensity).
             The result of keeping this distance between the two microphone arrays is
             that any leakage of the direct sounds into the ambient microphone do
             not dominate the desired ambient sounds going to that microphone
             array. This makes it possible to control the mix of direct and ambient
             elements, keeping phase and time delay effects to a minimum. Care must
             be taken not to position the ambient microphone array too far back,
             however, which may cause 'slap' effects when it is mixed with the direct

             Double MS recordings typically feature a spacious sound with good stereo
             imaging and definition. The ambience is more controllable than when
             using some other microphone techniques, and most importantly, the
             stereo image produced will collapse to mono without problems, the
             reasons for which have already been discussed.

             Another use for double MS techniques is in surround recording. Versions of
             5.1 stereo surround are currently being developed for music recording,
             based on double MS techniques, with an added sixth channel overhead.
             If you are able to make a double MS recording onto multi-track tape and
             play it back using a front and rear pair of monitor speakers, a very
             believable quad image will be produced, exhibiting quality and
             believability that is unobtainable using X-Y microphone techniques.

Suggested Microphones for MS Recording

             •A pair of AKG C-414's (switchable polar patterns).

             •A pair of Neumann U-87's (switchable polar patterns).

             •Shure SM69, dual capsule microphone (with switchable polar patterns for
             each capsule).

             •AKG C-422, dual capsule microphone (with switchable polar patterns for
             each capsule).
•Soundfield ST-250, multi capsule microphone (with switchable polar

To top