Ok, here's a weird one. Say you only had two mics and didn't have any money at the moment to buy any more. One is a hypercard, the other is a cardioid. You want to record in stereo, but your image will be all weird if you use the two mics in a more conventional stereo pattern.

So, you think to yourself, what if I set up with one mic on top of the other, both pointing in the same direction? Then when I get to my Compy, take the Card's signal, duplicate it, and flop the phase on one of them. Then bring the hypercard signal up the middle to fill in the phased information?

M/S-ish, right?

9 Answers 9


It won't work. I believe that your set up is missing a key ingredient which is a bi-directional microphone. The reason M/S matrix-ing works is related to the way the "positive" and "negative" sides of the bi-directional side mic interact with the mid's signal. Because both sides of the bi-directional mic's transducer are open, it is able to record a sort of "encoded" image of both the left and right fields. That is why the process of duplicating the side channel and inverting its phase is called "de-matrixing" or decoding.


not really. m/s technique relies on the fact the pick-up patterns between the two microphones overlap minimally. with what you're suggesting...not only will your resultant stereo-field be extremely narrow, but your going to have a lot of weird phase issue up the center of that stereo field...meaning that you're probably going to lose some of the signal that would have been there otherwise.


IS that to record ambiences or spot effects? If its for ambiences i'd put them back to back (180º) like 3" appart I've tried it after Charles Maynes suggested it in a forum thread and liked it to.

I like your idea though, should try it soon.


What you get here will be perceived as stereo, but it will be fake and suffer from an extreme phasing-problem in the right channel (providing it's the right channel you reverse, it normally is in M/S). What you have here is actually, except the phase-reversal part, something I've used on several occasions to make broken, phased-out, FUBAR sound effects. In mono.

The reason M/S actually works so well is that the Figure Of Eight-mikes are dually polarized, ie the membrane is positive on the front side and negative on the rear side. Then when you split up the Mid into two discrete channels with phase-reversal on the right one, left channel will enhance the positive signal in the mid-mic and attenuate the negative, and the right channel will do exactly the same thing, though opposite.

When you do not have an Eight to record the Mid's with, there will be no difference in phase between left and right, and after matrixing you will get a signal with the left signal having a slightly higher directivity than the Cardio though lower than the Hyper and about 1.5dB louder (depending on how well you matched 'em), the right channel will sound thing, phased-out and without a core. You can never replace the Eight in an M/S-recording, but you can actually do pretty much what you want with the Mid! I don't have any Figure Of Eight's right now since one of my Oktava's went belly up, but when I did I experimented a lot with everything between Omni's to Lobar's, and even another Eight with very nice results! Unlike in for example X/Y (personal favorite) or A-B and such, it doesn't really matter if the mics are matched or not, I've tried out everything from Oktava, Sennheiser, DPA, AKG, Shure, and several other high quality mics (I like to work with micings so I put my paws on everything I can get!), and even some rather bad one's from both ADK, Röde and Pulse (don't ask...) to mention a few (as I said, EVERYTHING I can get :-)) with different sounding but often greatly working results depending on what I want right then :-)


no - the point of the fig-8 in an MS pair is that one side is completely out of phase with the other.


I personally think discrete stereo sounds pretty good - try it out! Kind of like how Tim Prebble sometimes records ambiences - see his photos of how he records on his blog. If you space out the mics pretty far from each other I think it would work fine.

http://www.musicofsound.co.nz/blog/samoa-field-recording-03-night-and-day Correct me if I'm wrong, but he was recording a stereo ambience in the second to last photo, right? That's a pretty wide stereo and this technique should work fine with your 2 mics]

You're right that they may not do well set up in ORTF or X/Y, but I don't think you'll necessarily notice a difference in the wide discrete stereo field by having a hyper-cardioid and a regular cardioid unless you're recording things with very loud and distinct transient information like someone hammering at a construction site or something like that.

Also try out the M/S and let us know what you find - I'm curious if it would work!


I agree with @Miles & @Shaun, you need a bi-directional or to be more precise a microphone with a figure 8 pattern. If you invert or flip 1-channel of the cardioid or omni and summate with the original you will end up with +1-1=0, so all you would hear in your arrangement would be the signal from the hypercard.


What if, for the S, instead of using 1 figure 8 mic, you used two cardioids with opposite phase, summed together?

  • 2
    Personally, I feel like that's over-complicating. If you've got two cardioids, just go with some form of an X/Y or ORTF variant. You've already got the two channels for stereo, just record it that way. Also, there's a null point in a figure eight at 0 and 180 degrees. That helps reduce phase issues in M/S...there's no overlap. A cardioid only has a null pattern at 180, so there can be some overlap at 90 and 270 between to carioid mics...introducing possible phase issues. Commented Sep 6, 2011 at 1:40

i see this is from a while back but i was just trawling the net about this same idea. Had been playing around with faking stereo through m/s encoding and came to the same conclusion that it could be done with 2 cardioids.

Am i not right in thinking that it will add stereo image to the 'M' signal but be canceled out when played in mono providing you treat either side signal identically. To exaggerate this you could delay and drop some lows freq on the the 'S' signal in your DAW, or place it further back during recording. Adjust the delay to find a sweet spot where the phase relationships don't hollow out the 'M', if thats a problem.

The biggest down-side compared to using a 8 mic would be the minute differences that are picked up by each side (and these i assume add colour and prevent the sucking effect of closely or perfectly out of phase audio).

Let me know your thoughts on my method and i'll get back and let you know how it worked out in practice. :)

Oh btw.. i'm talking matched pair here.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.