Topic: 2 stereo mics > 2 channel recorder?

[yug du nord]

Crazy question that I've been wondering about...  Maybe even a ridiculous question... 

Technically, what would happen if 2 mid-side stereo mics were ran into an 1/8" stereo Y-adaptor (2 stereo inputs > 1 stereo output), which was run into a 2 channel recorder?  That is...  one mid-side stereo mic in the right channel, and one mid-side stereo mic in the left channel.  Not asking why one would do this...  but asking what would happen...  what would end up on each channel?  I have my opinion what happens, but without factual knowledge...  I'm only guessing!!!  I can't seem to wrap my brain around this concept!!!
                                                                                                 

[Sami Ollas]

You'd still get a two channel recording where each channel would have a merged audio of what the stereo mic picked up for the said channel.

[yug du nord]

You'd still get a two channel recording where each channel would have a merged audio of what the stereo mic picked up for the said channel.

That's my exact thought...  but is that correct???

[Sami Ollas]

I can't see any other possible solution for that really. Since the recorder only has 2 channels, there's no possible way for you to get 4 channels with that even if you'd use a splitter like that. All it'd do is pick up the sound with all 4 mics and combine the stereo mics into the left and right channel. Essentially what you'd get is a matrix of those two microphones.

[yug du nord]

Any chance that since they are both stereo mics (R/L channels), that only the left channel from each mic (mono channel) would be sent to the recorder?  So it would end up recording:

:right channel of recorder = left(mono) channel of stereo mic (right mic)
:left channel of recorder  = left(mono) channel of stereo mic (left mic)

[Sami Ollas]

That depends entirely on how the splitter is made. For example a headphone splitter that allows you to listen to iPod (as an example) with two headphone sets, still allows both listeners to listen to it in stereo. If the splitter is similar, the wiring would be made in such way that they're combined. If the input jacks on the splitter are mono, it would do the same thing since the right and left connectors would both be connected to the same spot, the ground on the other hand is on the exact same spot on a mono miniplug than it is on the stereo plug.

[yug du nord]

+T to ya Borlag!!!  I am talking a stereo Y-splitter.  So, you and I are thinking alike on this matter...  what you're saying makes complete sense to me...  I figured that is how it would work, so it's nice to get some confirmation...  but I'd be interested in getting a 3rd opinion...  anyone??? 

[attheshow]

Makes sense to me. I just ran a simple test that I think proves it... Plugged a splitter into my headphone chack and plugged my headphones into either "side". I get music in both ears. Presumably, this would work the oppposite way as well... mixing both channels into one "side" when going from a pair of stereo mics in.

[DSatz]

uncleyug, you'd be placing the signals from two low-impedance microphones in parallel on each channel of the recorder. That's a no-no in general, because low-impedance microphones are designed to work with a load impedance of 1000 Ohms or more. You'd be placing each microphone "across" the other, so they'd each "see" a 30-to-200-Ohm load. That will overload most condenser microphones and cause severe distortion. Dynamic mikes wouldn't overload as such from this, but their frequency response would be affected considerably.

That's why Mackie gave mankind mike mixers (say that five times quickly!).

--best regards

P.S.: If you combine (mix together) the signals from an M/S pair of microphones, you get the left channel of a stereo recording. So if you mixed (rather than simply paralleling) one pair of M+S signals and fed that to the left channel of your recorder, and also did the same thing with another M/S microphone pair and fed that result to the right channel of your recorder, you'd end up recording the left channel signals from each of your two M/S pairs. That recording would probably have a weird balance, depending on where you placed your two pairs of mikes.

[yug du nord]

DSatz...  thanks for the input!  Any chance you could "dumb it down" for me?  About the impedance and ohms?  The mics in question have an output impedance of 1kilo ohm (=1000 ohms) +/-20% unbalanced. 

I'm asking this because I ran a pair of Sony ECM-MS907 stereo mics in a DIN(not exact) configuration into a minidisc recorder for a few years.  I never had a distortion problem though.  Vocals are occasionally thin...  but I think that's a result of "cheap" mics.  I am just now starting to go through and transfer some of those recordings.  I would also run a single ECM-MS907 for stealth situations occasionally.  I tend to think that using the pair of mics sounds better than the single mic.  So, I've been wondering...  why that would be?  Wav files look good, no imbalance issues, no distortion with either method.  I pulled decent recordings both ways, so I always stuck to it!  Why change a decent thing?  But now, I'm a bit wiser and more curious as to what I was actually doing!  Any other thoughts?  Thanks!!! 

edit:  these are electret condenser mics.

[DSatz]

uncleyug, maybe I don't understand exactly. I don't know that Sony microphone, for one thing. Does it have M and S outputs rather than left and right? That would actually be rather remarkable in a consumer product, since most consumers don't have M/S matrixes.

This makes me suspect that it's actually just a conventional stereo microphone (i.e. one with left and right signal outputs), which just happens to use M/S as its way of generating a pair of left and right signals. Is that the case?

--best regards

[yug du nord]

This makes me suspect that it's actually just a conventional stereo microphone (i.e. one with left and right signal outputs), which just happens to use M/S as its way of generating a pair of left and right signals. Is that the case?

--best regards

Stock photo at bottom...

Its description from Sony is a:  "One-Point Stereo (employing the Mid-side stereo system) electret condenser microphone" 

You are correct!!!  It is a stereo mic but has the "Mid-Side" stereo system...  not the conventional left/right config.  But it is a standard L/R mini-plug termination.  It looks like a small condenser mic.  On the mic, there is the main "Mid" capsule, and just behind that (on the mic body) is a "Side" capsule located on each side of the body...  one on the right side and one on the left side of the body.  If that makes any sense???  To my understanding, the "Mid" capsule (considered mono) is mixed(-/+) with the left capsule to produce the left channel signal...  the "Mid" capsule is also mixed(-/+) with the right capsule to produce the right channel signal.  More technical description below.

Here's a description from Sony:

:One-Point Stereo design
:Mid/Side (MS) capsules for natural stereo panorama; the Mid capsule picks up monophonic sound while the Side capsule picks up left/right difference sound; subtracting and adding the two capsule signals yield separate Left and Right channels - also permit electronic adjustment of pickup angle
Mid/Side (MS) switch selects pickup angle between left and right channel; choose 90° for a single voice or instrument or 120° to pick up many voices and instruments, arranged across the stage

>>>+T...  Thanks again!!!<<<

:edited for photo

[DSatz]

uncleyug, as I suspected, this is an ordinary coincident stereo microphone with left and right outputs. Internally it uses M/S to generate those left and right outputs, and that's certainly an interesting bit of applied technology, but it means that in the end, you simply have a coincident pair of directional microphones there.

Compare this with microphones that generate an M and an S signal directly (the S mike always has a figure-8 pattern), which you can then dematrix into left and right (choosing your own proportions of those two signals, as well as the pattern of the M microphone), or you can record M and S and then dematrix later on, at home while listening over your favorite loudspeakers so that you can choose the proportions to match. That's M/S recording.

You probably know this already, but any coincident pair of microphones can be matrixed into M/S and vice versa. But the whole thing about mids and sides, you can forget as far as your original question is concerned; you have two coincident stereo microphones of the same make and model, is all.

OK. If you have a pair of coincident stereo microphones aimed a few inches apart, with some angle between them, more or less as if they were ordinary, single cardioid microphones in an ORTF or similar array, and if the outputs of those microphones are paralleled, what you have is rather difficult to characterize because it is very irregular geometrically, but it is symmetrical on a left/right basis, so in that sense it's not unsuitable for stereo playback. Still, whenever you mix together the signals from two microphones that are only a few inches apart, what you'll get is reinforcement of low frequencies and then, starting at some upper midrange frequency (that's a function of the exact distance between the microphones) you'll get alternate bands of cancellation and reinforcement--what's known as a "comb filter" effect.

However, your two microphones per channel aren't aimed in the same direction, so their cardioid pattern would tend, I think, to reduce the comb filtering effect somewhat.

All in all, I really wouldn't recommend doing what you did, for all the reasons I mentioned earlier (loss of sensitivity and headroom) and because the resulting pickup pattern from the two mikes combined per channel would be irregular especially at high frequencies.

--best regards

[yug du nord]

Thank you...  one hell of an answer!!!  Nice of you to step-up with a great response to a strange question about an inexpensive mic!!!  Your replies here have made me do some homework...  thanks man...  a good lesson has been learned today...  in many subjects!!!   

But...  as to my original question...  what got recorded on the left channel of the recorder, and what got recorded on the right channel of the recorder?

First is a quote from Sony:  "Mid/Side (MS) capsules for natural stereo panorama; the Mid capsule picks up monophonic sound while the Side capsule picks up left/right difference sound; subtracting and adding the two capsule signals yield separate Left and Right channels - also permit electronic adjustment of pickup angle." 

So......  do ya think it is the full stereo (L/R) signal which is produced by the Mid/Side config. of the left M/S mic that is sent to the left channel of the recorder?  And also the full stereo (L/R) signal which is produced by the Mid/Side config. of the right M/S mic that is sent to the right channel of the recorder?  Or...

Do ya think that the left M/S mic is sending a mono signal (only from the left M capsule) to the left channel of the recorder?  And also the right M/S mic is sending a mono signal (only from the right M capsule) to the right channel of the recorder?

Any more thoughts on this?  I think the answer is just around the bend...   

I'll ask ya also...  since I got ya here...  if you could clarify the M/S configuration?  M+S = left channel and M-S = right channel...  correct?  What exactly does that mean?  M+S = left channel... I understand (I think).  But, M-S = right channel... I don't understand.  Any help on that one?

                                                                                                   

[DSatz]

uncleyug, the basic way M/S recording works is that there is one forward-facing (mid or "M") microphone of any pattern you like, and one side-facing (side or "S") microphone which has a bidirectional (figure-8) pattern.

The two microphones are set up so that the centers of their diaphragms are aligned one above the other--so as far as the horizontal plane is concerned, they're at the same point in space (= "coincident"). Any sound which reaches them both will reach them both at the exact same time, because they're both the same distance from anywhere a sound could be coming from, in the horizontal plane.

But one basic fact about a figure-8 pattern is that its two lobes have opposite signal polarity--they're "180 degrees out of phase" with each other. So if a sound is coming from, say, 45 degrees to the left of center, it will of course reach the M and S microphones at the same instant--and since (in a normal setup) the left lobe of the figure-8 is "in phase" with the forward-facing M microphone, the signals generated by M and S will have the same polarity as each other. Compare this with a sound that originates from, say, 45 degrees to the right of center--it will reach both microphones at the same instant, but it will be picked up by the lobe of the "S" microphone that has oppposite polarity, so the M and S signals will have opposing polarity (they'll be "180 degrees out of phase" with each other).

OK. Now, you were exactly right when you quoted the formulas "M + S = L" and "M - S = R," and you wanted to know what they mean. Try them out on the example I just gave. In the first instance, with the signal that was coming from 45 degrees to the left of center, M + S = L means that to obtain the left channel output at that moment, the matrix is adding two signals that are "in phase" with each other, which will give you back 2 x the same signal that you started with; meanwhile the right channel output (M - S) means that the matrix is subtracting the same two signals from each other, which leaves 0 (or something close to it). And that recreates exactly what we started with--a signal in only the left channel.

If you follow this through for the other example (where the sound originated from 45 degrees to the right of center) you'll get the same result in reverse: approximately 0 output in the left channel, while 2 x the original signal will show up in the right channel output. So that's what those matrix equations mean. When you have multiple sound sources all at the same time, their signals are superimposed but the polarity situation of the figure-8 microphone (it's really just a cosine function, if you took trig) sorts it out, so to speak.

The thing is, the matrix is built into your microphones (it can be implemented very simply using op amps) rather than being an outboard component. So if it makes things simpler--and I think it might--you could separate "what happens when you parallel two stereo microphones" from "how does M/S work?" completely. You paralleled the outputs of two stereo microphones, and those outputs weren't M/S signals; they were perfectly conventional left/right stereo signals which happened to come from microphones that (internally) are constructed on the M/S principle. But the same signals might equally well have come from coincident stereo microphones that didn't use the M/S approach at all.

Is this starting to narrow down the size of the mystery?

--best regards