roon wrote:Ignorant to the MS matrix, i have done some research. I need to understand how it works in the analogue world.
Let?s assume for sake of hypothesis that we are recording a sine from a mono point source. We chose 2 microphones. A direct or cardioid mic pointed directly at the source (MID - M). And a figure 8 mic on a 90 degree axis point at the source of the primary reflections (SIDE).
The direct mic will record the exact sound, and the off axis figure 8 mic will record reflections. As it is a diaphragm mic, the positive side of the wave will be the left reflection and the negative amplitudes will be the right reflection. Please feel free to correct my interpretation.
This is probably the most common error of understanding that people have when they begin to learn about M/S. The side mic is a single, monaural microphone - it has a single output, a single signal - it's mono. Because it's a bi-directional mic, it JUST HAPPENS to have a pickup pattern comprised of two lobes of sensitivity - in this case, the mic is most sensitive to sounds arriving from the left and from the right.
Here's where it gets confusing: one mic signal/channel, comprised of two lobes of sensitivity. Even more confusing is the phase relationship between the two lobes, but if you think about the microphone's structure as having a diaphragm that is open to the air on both sides, and is made in such a way that diaphragm motion in one direction produces a positive voltage at the output, whereas diaphragm motion in the other direction produces negative voltage at the output. In our M/S setup, if there was a pressure positive event (like an impulse or a balloon popping) happening on the left side, in the first instant, the diaphragm would move away from that, towards the right. Let's say that we've set our mic so that when this happens, it produces a positive voltage at output (normally referred to as the "front" of the bidirectional mic). If an similar pressure positive event happened to the right of our M/S array, the S mic's diaphragm would have to move towards the left, and following how the mic is set up, this would produce a negative voltage at output. Perhaps you can also see that if you take our balloon and pop it somewhere in the same plane as the surface of the microphone, the pressure would reach both sides of the diaphragm at the same time and at the same intensity, so the diaphragm would not move - it's the most insensitive region of the mic's pickup pattern, or the null.
Remember that I'm talking about the instant that the impulse's wave front reaches the microphone diaphragm. At any point in time afterward, different conditions will apply, but the underlying principles will still apply - I'm just trying to simplify things for the sake of the explanation.
The next stage is the route the signals from both our M and S microphones into a mixer. In the good ol' analogue days, you'd actually use a splitter wire attached to the output of the S mic - one input, with two outputs. Remember it's still just a single signal, but now we have two identical copies. Now, we take one of those two outputs and invert the signal polarity (most easily done by swapping the wired connections in the XLR connector, exchanging pins 2 and 3). Now we have one signal, with two outputs, and one of those is polarity inverted. So now you have THREE inputs for the mixer: One from the M mic which you then pan to the centre, and the S signal (not inverted) panned hard to the left, and the inverted S signal panned hard to the right. You'd match the levels of the two S signals and then you'd vary the mix between how much of the M and how much of the Ss you'd like - more M and less Ss makes for more mono, and more Ss and less M making it wider and wider stereo.
The key here is to remember the function of the panpot. The panpot for the M signal's channel is centre panned - meaning it is sending an equal intensity signal to both the left speaker and to the right - so yes, it's splitting the signal and sending it to two destinations, but in this case both splits are equal, and IN PHASE. This give the listener the impression that the sound is emanating from a point in the middle of the speaker array.
Because we've split our S signal before even getting into the mixer, we're dealing with two channels - one with it's output to the speakers panned left and the other channel panned right. But really, with a more clever mixer architecture, like what you can find in most modern digital mixers - why not simply put the polarity inversion inside the mixer - say, just before the panning. Just as we saw with the M signal, we have one signal going to two speakers, with each split being equal and in phase. For the S mic, let's just do the same thing with one exception - let's invert the polarity of the signal running to the right speaker, but still send an equal intensity of signal to both.
You laid out the math in your original post, but perhaps the math doesn't really show what's going on, so let's examine things closely. A key item to remember is that two signals that are equal in intensity and IN PHASE, then they will combine constructively, and the result will be a louder output for that signal. On the other hand, if you combine two identical signals, with identical intensity, but with one of them phase inverted, then they will combine destructively, and will cancel each other out, resulting in no output at the speakers.
So....
We have our M/S mic array set up pointing at a stage. An event occurs towards the audience's (and the mic's) left. When the wave front of the initial impulse reaches the diaphragm of the M mic, it will move backwards (towards the back of the hall), producing a positive voltage at its output. At the same instant, the diaphragm of the S mic will move from the left to the right, also producing a positive voltage at its output. At the mixer, the M signal will be sent equally to both the left and right speakers - as will the output of the S mic, with the all important distinction that the signal sent from the S mic to the left speaker will be in phase with the M mic's signal, and therefore will combine CONstructively, making a louder output from the left speaker - AND - the signal from the S mic will combine with the M signal DEstructively on the mixer's output to the right speaker, resulting in a quieter signal output from the right speaker. Put a human with good hearing in front of those two speakers, and they will experience the output of the two speakers as emanating from a single, virtual or "phantom" position, somewhere between the two speakers, but mostly towards the left.
roon wrote:Some simple addition of ordinates of these amplitudes can isolate left and right and middle channels. This is MS decoding.
No, MS decoding doesn't isolate the three channels (but only two mic signals) - it combines them to create two signals, one for the left speaker, one for the right.
roon wrote:Centre M-S decoding requires a sum-and-difference matrix, where you add one side signal to the mid signal to get the sum, and subtract the other side signal from the mid signal to get the difference.
No. In common use, the terms "sum" and "difference" refer to those signals coming out of the speakers. The "Sum" refers to those parts of the signals that are the same intensity and polarity from both sides. The "Difference" refers to those parts of the signals that are NOT the same - that's all those terms mean in this instance.
roon wrote:To do this, one of the side channels is shifted 180 degrees in phase: when the polarity-shifted signal is added to the mid signal, you get the difference between the two signals. Using M for mid, S for side, and -S for the polarity-shifted side,
M + S = left channel
M + (-S) = right channel
Of course the mid is already isolated in a track of its own.
You've stated the math correctly, but I can't help but feel that you have misinterpreted what's actually going on - I hope my long winded explanation has helped with that.
roon wrote:SO how does this translate to the micstasy? Do adjacent channels become the 2 microphone channels. MID mic on channel 1 and SIDE mic on channel 2. SO if MS is selected on both channels and phase invert and MS decode is on in channel 2 i would get the right hand side? Which output does the result come from? 1 or 2?
As all this applies to the Micstacy, I don't really know, as I've never even seen one, BUT, I expect that it has some sort of either hardware or software mixer associated with it, that can route signals from its various inputs to various outputs. If it's truly clever, it might even have an M/S decoder built right into the mixer, as described above. But you don't really need a Micstacy to use M/S - the principles remain the same no matter what mixer you're using - even if you have to resort to the old fashioned idea of using a splitter wire for the S signal and three channels on the mixer.
I think the confusion arises from the fact that you have two signals coming into from the mics - the M and the S, and you have two signals going out to the speakers - the L and the R - but they're not the same signals at all. M and S are not the same as L and R, but you can make L and R out of M and S, and viceversa, through some clever combining and inverting of signals along the way. The benefit is that doing things this way allows you to, in effect, MOVE the microphones to create a wider or narrower stereo image, without having to leave the mix position to move microphones - you can just adjust things to your preference while listening to the result.
M/S is used in many other applications as well - FM radio is broadcast in M and S, and then turned into L and R in your radio receiver. LP disks have the audio signal cut into the grooves so that the vertical motion of the stylus produces the M signal, and the side to side motion the S, which in turn get turned into L and R in your amplifier. There are lots of reasons they decided to do things this way, but perhaps this post is getting a little long already...
roon wrote:Perhaps i have even researched the wrong topic and this MS decoding is something completely different.
PS I am sorry for asking so many questions. I do not want to come across as a LOUDMOUTH. Indeed , the opposite is true.
with all due respects,
roon.
No problems - I hope I've cleared up some of the confusion for you, and for anyone else interested in this topic.
Frank Lockwood
https://LockwoodARS.com
Fireface 800, Firmware 2.77
Drivers: Win10, 3.125; Mac, 3.36
