This post comes from guest contributor, Charles Maynes. Charles is a talented Sound Designer, Editor, Recordist and all around decent fella.
I’ve also added a few footnotes that might be helpful -Michael
As a child of the Cold War, I always loved the film Dr Strangelove, and now can mark off my bucket list my misuse of one of its most humorous tag-lines: “so with that said, in we go …”
I am going to need to give a little disclaimer, this little piece is an opinion piece, with little science attached to it. There is plenty of great writing out there as to the principles of Mid Side microphone techniques[1], and it isn’t my intention to add to the serious academic tone of those. There is no “answer” at the end of this. It is just a batch of information that will empower you to decide whether this is a good fit for the way you might want to do things.
As a sound designers, sound editors and sound effects recordists, we have all sorts of techniques and practices at our disposal. We use high sample rates and different bit depths. We also work in mono, stereo and multichannel formats. Usually, we are constrained by one condition or another, and many times we are simply constrained by time itself. In the field and in the studio, working with stereo signals we usually are most comfortable with Left, and Right channel audio. The reasons for this are pretty simple: most people with any sort of audio skills are usually very comfortable working in that format. And to be stating the obvious, our reproduction systems pretty much demand that. When we look at more esoteric formats like M/S, double M/S, B-Format and discreet surround, as the Holophone class of microphones provide, we are quickly swimming in a sea of options which may or may not make our creative work more efficient. Since I am discussing the basic Mid Side technique, I will hit the pros and cons as bullet points, which I will address in greater detail:
Mid Side “Pros”
- A real mono center channel- this is pretty important, and quite cool.
- Stereo width that is adjustable after the fact.
- The same channel requirements as other stereo mic’ing setups
- single point microphone placement- which is more convenient, and can be hand held with relative ease and comfort.
- The ability to use different pattern mics for special purposes and still have a stereo compatible result.
- The ability to use processing on the Mid or side mics independently.
Mid Side “Cons”
- PITA file management
- PITA auditioning in library programs
- PITA use in mixing
- PITA use by team members not versed in M/S work flows.
I am sure there may be other pro’s and con’s, but those are my pet favorites. Now to dig into the list!
On the “Pro” side.
- A real mono center channel- this is pretty important, and quite cool. There is NOTHING better than a true mono recording of an effect if you need it. Mixing stereo down to mono is sometimes workable, but in most cases we just chuck one side and move along. Is that good enough? Well in most cases it may be, but its not as good as a nice mono recording in the first place. The one thing to recognize is that all the other stereo microphone techniques are really based on the idea of recording the space around the subject, so all the mics are off axis to one degree or another- in M/S, that is not the case. Our Mid mc is pointed directly at the subject.
- Stereo width that is adjustable after the fact. That pretty much states fact. We can add or subtract the side channel at will to determine the degree of space we like. The added benefit for this is that our side mic is an off-axis angle which might actually be useful on its own. One thing that is possible via plugins is that we “could” convert X/Y or ORTF recordings to M/S via plugins, but we will not end up with the same flexibility due to the fact that we must use matched microphones for those other techniques (or at least “should” for somewhat convincing stereo).
- The same channel requirements as other stereo mic’ing setups. M/S is a 2 channel format, and where things get interesting is if we add a third channel with a rear firing mic in order to achieve double M/S for a 4 channel result, and since we have a discreet mid mic, that could also become the center channel for a 5 channel recording. Pretty cool stuff!
- single point microphone placement- which is more convenient, and can be hand held with relative ease and comfort. Again, this is pretty obvious, we can do an MS mic package in a single blimp with not too much trouble- something that is more of a hassle with ORTF or spaced omni setups. It also means it requires only one mic stand to carry for the stereo setup….
- The ability to use different pattern mics for special purposes and still have a stereo compatible result. This is where things get interesting a cool. Since you have two mics, you can use whatever you like for your M mic[2] (pretty much) you could use everything from a cardiod to a long shotgun and still have a variable width stereo signal- other things though that are less standard is say, using a core M/S package with a hyper cardiod and figure of 8, and running a second forward firing mic that is a short shotgun, or even a dynamic. There’s lots of cool potential there.
- The ability to use processing on the Mid or side mics independently. This is one of the most compelling things the M/S format offers. You can change the stereo image by eq-ing the mid and side mics independently. Once you try this, I am sure a good deal of people will be very much into the results. This provides tonal shaping that is not easily available in other stereo sources.
Now to the negatives….
- PITA file management M/S files live in their own universe as far as editing programs go- if we use M/S clips in our editorial work, we need to have the ability yo use plugin matrix solutions to yield our expected stereo output. This requires any M/S tracks to have those plugins as a part of their channel strips, and though simple, they do require system resources. Some plugins are very simple like Wave’s MS Matrix which basically has a set balance level which is not adjustable from inside the plugin, so in order to use that effectively, one would need to run a double mono trim plugin in front of it to be able to control the mid to side balance. It is not a deal breaker, but a PITA. Other plugins like HOFA’s level control plugin (which is free) have better control, but still need to automate if the channel is using both M/S and standard stereo clips in it.
- PITA auditioning in library programs The problem is basically the same as above. Our librarian tools need to be able to decode M/S and hopefully do it without too much effort[3].
- PITA use in mixing The same core problem as in #1. M/S is not dub stage friendly, and requires forethought by the mixer to effectively use it. I cannot think of many mixers who look forward to getting M/S tracks instead of regular stereo ones.
- PITA use by team members not versed in M/S work flows. The issue here is that if M/S files are a part of an editorial workflow, it will take editors extra time and effort to deal with them. They largely don’t want to spend this time since it takes away from editorial, and they probably aren’t going to be enthused about it. This is mainly due to the anxiety of whether the mixer wants the tracks instead of straight left/right stereo. Making lives hard for mixers usually isn’t a great career move either.
Conclusions
I have to say that I have moved from being anti-M/S (due to the reasons above) since getting my Oktava setup up and running. I have worked with a few M/S rigs and found them not particularly compelling, but the Oktavas are making me very happy about what can be done with the technique.
I love mics and I love stereo. I regularly use lots of neat and lovely stereo mics. The Sanken CSS5 is probably my “desert Island” mic. The SASS gives a type of binaural effect which can be very nice, and lastly my old, but still appreciated Shure VP–88, which sort of steered me away from M/S in the first place.
Now you are probably thinking that with the most popular M/S setups being either the Sennheiser MKH range, or the Schoeps which are certainly an industry standard, how did I end up with the Oktava rig? Well, I have to say it was an odd twist of fate for the most part. I read Tim Nielsen’s great articles on Designing Sound and focused my attention back on the M/S world, and then some collaborators I know, Zach Seivers and Justin Davies quizzed me on the notion of M/S recording. They were both looking to get Schoeps rigs for a film they were starting on and I spent some time going over the pros and cons, much of which I have already mentioned here, and let them sort the idea out for themselves. As it turned out, they both thought that the added effort was worth it, which pushed me towards putting together another MS rig myself. The Shure VP88, is an M/S mic with a built in decoder to provide a stereo or discrete output, but for whatever reason, I really never got great stuff from it (with the exception of discovering it to be an awesome mic for drum overhead). I thought I needed to rethink things, and look into options that worked with mics I already had.
My favorite small diaphragm condensers I had were the RODE NT–6’s and the Oktava MK 012’s. Each performed well, but largely were utility mics for me so as I continued considering different options, I came across an announcement from Oktava that they were doing a figure of 8 adaptor for the MK012’s. I had 3 of them already, and a good collection of capsules I thought this would be a not too expensive way to go. After getting it finally from the distributor, and rigging it up sort of awkwardly in a RODE blimp I noticed a lot of things which I didn’t really like. The mics were awkward to get a stable mount for using bongo ties and cut pieces of foam. This was remedied cheaply by the Rycote stack clips, which secured the two mics perfectly. Now the second issue seemed a little more difficult to deal with, and that was running two mic cables into the blimp and not having an ocean of bumps and other distortions if I dared move the blimp. Obviously, this was unacceptable.
I remembered another friend, Tom Hambleton mentioned the Rycote Connbox. I looked it up, and was disappointed that it seemed to be a simple device with a pretty big price tag for what it seemed to be. The one feature it had that I DID love was that it allowed me to use 5 conductor mic cables as I did with the Sanken and Shure mics. I had recording gig to cover it and I finally popped for it. The results with the Connbox were really unbelievably good. All the cable noise was pretty much gone, and it was very clean and could be easily moved. In then end it is an amazing value, which I have no reservations saying is an awesome tool.
One thing that is very important to your sanity is carefully aligning the Mid and Side mics in your shock mount. I started with my Mid mic a little bit forward of the side capsules and had difficult in mastering figuring out the right amount of delay that was need to produce a correct stereo image after the fact, but it was certainly a good learning experience.
So with that, here are some sound example of the mics for you to peruse, and hopefully you can make a good decision if M/S is right for you.
M and S files:
Matrixed Files:
- There is a wonderful paper written by Wes Dooley and Ron Streicher that lays out some detailed info here (PDF) – MR ↩
- pages 4–9 of Wes Dooley and Ron Streicher’s above paper layout the impact of the using different pickup patterns – MR ↩
- Soundminer can matrix files during previews if you put “M/S” in the channel layout field. It is a handy way to quickly preview your work if you want to hear it matrixed. -MR ↩
Always informative to read your thoughts, Charles; thank you. I may follow suit on that Octava rig you’ve got going there. RIght now when I’m recording ambient sound, I double up with the binaurals and the MKH418s (which you know I’ve been fond of for many years). Though I’m curious to experiment with a 3rd stereo perspective. I have a recording trip coming up and I was considering the CSS5 (based on your numerous endorsements over the years) as a 3rd option, but I may try out the Oktavas first before making that investment.
I use M/S a great deal (Sennh. MKH-8040 and MKH-30)
Typically I record thru matrix as stereo, via sound devices 702T built in matrix. And thus avoid post complications and confusions.
Sum L+R to mono (or simply pan center), and voila, the side mic sums to zero and you have beautiful, on-axis mono. If you need wider, plugins such as Waves S1 or Flux Stereo Tool are very convenient.
Excellent article Charles. I just recently sprung for an MS setup myself and I have to say that I absolutely love it and will probably not go back to XY again.
For all the hassles you describe above I still think it is a worthwhile recording method. I still remember the first time I brought up the side tracks and combined them with the mid mic… I was blown away by the realism and spatialization… something I never got with XY.
I purchased a Schoeps rig and I love it. Combined with a 722 you cant go wrong.
[…] designer and recordist Charles Maynes has written a great nuts-and-bols article about mid-side recording, posted on the field sepulchra blog a few days ago. The theory and practice of Mid-Side recording […]
Great article Charles! I was just asking around about MS options, so nice to read a long post by you about it!
Great article with some great examples to listen to. Alan Dower Blumlein’s MS technique is even more under used than his Crossed-8s and both give a great sense of spaciousness due to contradictory phase issues. All the more remarkable given that there is essentially no time of arrival differences at the capsules.
This conundrum led the famous John Watkinson to have the strap-line MS-Maybe Stereo.
Double MS is a fantastic solution for surround recording. (Though it requires care, M in C with S in LR does note decode well except in the air which is
wrong but interesting)
http://www.cinesonics.pt/Locmat/Manuais/Schoeps_DoubleMS_Paper_small.pdf
Thanks for the article – it’s always good to read what others do in terms of practical issues with cables, suspension and such.
I’m very much a DIY-guy (read: skint), but that Rycote stacking clip just makes sense. The Connbox triggers my tinker-finger though – possible to DIY?
Do you find the Oktava’s too noisy for quieter subjects like nature recordings?
Really nice to get a great recordist’s endorsement for a great technique. Beautiful article Charles.
One of the interesting features of MS is that it decodes surprisingly well into LR and from LR into Dolby Stereo. C=C, L&R=L&R, S=rear. If you widen the stereo image too much, though, the L&R will be pulled into the S too much. Using a Pro Logic decoder plug in it is possible to brain LCRS decoded sounds from a single MS setup. Use the MS decoder’s width control to adjust S balance along with LR width. Pro Logic II gives split surrounds. Mono up to 5.0: Nice to have that available from a single two channel recording!
I love the sound of MS decoded into discrete LCR. We used that, or emulated the sound of it in Cloverfield. Inspired by the work done by Gus Van Sant, Felix Andrew and Leslie Shatz on Elephant, whose production track was recorded in MS and decoded into LCR, augmented by lavs only where needed for clarity. There’s a smearing of the image as in nature, completely different from a mono panned sound, and different from an LR recording toed in to the center.
Charles, Do you get satisfactory results using unmodded Oktavas or modded?
The Oktava’s are pretty clean- though it seems there might be some variation from the individual mics= I will say I think they are sounding better in M/S vs my usual usage with them in X/Y or ORTF setups- I am using usually a Hyper Cardiod Mid capsule with the standard Cardiod capsules on the side mic- the one thing that is sort of interesting is the spacing on on tne side address mic which is definitely interesting. All the example fils were recorded with that setup though, and I think they have a rather nice timbre to them… Doug! thanks for your awesome comments too!
Hello Charles,
First I would like to thank for your insights a practical approach on this subject. I’m a MS fan and use it all the time, so I thought I could contribute with some thoughts. Regarding the file management it is not very difficult to handle if you decode your MS signals to stereo while recording. This is the great thing about MS, you can record it in stereo (L and R channels) and through does files into any DAW as regular stereo files, like you would do with any other stereo recording. Later if you wish to balance the level between Mid and Side signals, you just convert the LR in to MS again (M=L+R and S=L-R). There is a great plugin out there which is free that helps you do all of this, the Voxengo MSED.
http://www.voxengo.com/product/msed/
[…] in doing a blog post for his Rabbit Ears Audio, and that was essentially the fodder for my “How I learned to Love M/S.” With that, doing the blog post, I thought, “Well, you know, it’d probably be pretty […]
Hello Charles,
thank you for this nice review about M/S stereo techniquie. A thing what definitely shold be mentioned, that in comparison to X/Y, the use of M/S gives you the possibility to use a omni-directional microphone (pressure receiver) which sounds different (=better, more natural) to Cardioid microphones (gradient receivers)! One of the main reasons, why Jürg Jecklin invented the Jecklin Disc at that time..
Best, Wolfram