Stereo has added new dimensions to our recordings, an effect that is difficult or impossible to achieve in monaural recordings. New immersive formats, such as Dolby Atmos, add even more dimensionality to music.
You can achieve a kind of stereo by panning mono recordings of various instruments into the stereo field, but that is not what this episode is about. I talk about how to use a pair of microphones to pick up sound.
I provide a very brief history of stereo to get started, and some simple concepts of microphone pickup patterns and why they are important in stereo recording. Then I delve into some of the more popular approaches to stereo recording, including spaced-omni mics, X-Y and its variants, M-S, Blumlein, and binaural recording using an artificial head.
Mono compatibility is less of an issue than in the past, but still should be considered. I compare the various techniques and what happens when the two channels are combined in mono.
Some other techniques I have used over the years include mis-matched mics in stereo, and what I call “incidental stereo,” which could be from bleed between various instruments and mics, or just recording an unused nearby mic in the studio for possible combination with a main pickup.
I conclude with my impression of each technique, which, of course is entirely subjective and will undoubtedly be different than what you find.
This topic was suggested by listener Bill Sallak. If you have any suggestions for topics, I’d like to hear them!
Thanks for your comments and suggestions. And thanks for listening, rating, and subscribing to this podcast. It is available on over 30 podcasting platforms.
I’m Doug Fearn and this is My Take On Music Recording
Stereo recording became practical in the 1950s with the widespread use of tape machines. Over the next couple of decades, various stereo mic’ing approaches were developed. None of them is perfect, and none of them is ideal for every situation. But stereo added a literal new dimension to recording, and improved the listeners’ experience.
New immersive audio formats, like Dolby Atmos, give us even more opportunities to create interesting sound stages. But I am going to limit my discussion to traditional, two-channel stereo. Keep in mind that these techniques can also be used in any immersive format to great advantage.
In this episode, I’ll discuss the various techniques, their advantages and disadvantages, where each is most useful, and my experience with using them.
We often think of “stereo” as panning of individual monaural tracks into the soundstage. There’s nothing wrong with that, and it certainly makes life easier for the mixing engineer. But that’s not the kind of stereo I am talking about here.
Our discussion is about picking up sound as a stereo image in itself. You probably already use this approach for recording drums or piano. Actually, most drum recording is a hybrid of mono tracks, panned where you want them, plus an overall stereo image from the overhead mics and/or from room mics. A stereo piano recording adds width to the instrument and provides spatial separation of the notes on the keyboard. This discussion applies to those types of pickup, too.
But first, it’s helpful to start with the various microphone patterns and how they are used in stereo recording.
The most basic pickup pattern is omni-directional. In an ideal omni mic, the mic itself is a single point and sound is picked up perfectly equally from all directions. Not just around the horizontal plane of the mic, but also above and below, and everything in between.
Such a microphone does not exist. All omni mics in the real world are only an approximation of the pattern described; they are not “isotropic” in engineering terms. All exhibit some directionality in various planes around the mic. But most are pretty good at uniform frequency response through most of the horizontal plane around the mic.
Why would you use an omni mic? Well, it is easier to design an omni mic with flat frequency response than it is with a directional mic. And the frequency response is usually consistent throughout the pickup pattern.
Next, we have the bi-directional mic. This pattern is also called “figure-8,” due to its shape. Ribbon mics are natively bi-directional, but multipattern condenser mics usually have a bi-directional pattern option. Bi-directional mics pick up sound in front of the mic and in back of the mic equally. The sides of the mic have remarkably deep nulls, where nothing is picked up. And well-designed bi-directional ribbon microphones have uniform frequency response to sounds arriving from any direction. That’s important for pleasing stereo use.
The third basic type of microphone pattern is uni-directional, also sometimes called cardioid due to its heart-shaped pickup pattern. For this discussion, I will refer to them as “directional” mics. But keep in mind that there are many variations of this theme, including hyper-cardioid, wide-angle cardioid, or shotgun mics with a very narrow pickup pattern.
It's difficult to design directional mics with uniform frequency response in all directions. Most of them have at least some weird off-axis frequency response. This can result in bad-sounding audio from bleed from an off-axis source.
Another thing to remember is that most directional mics have their minimum sensitivity, or null, at some direction that is usually not at the back of the mic. Typically, directional mics will have nulls at about 135 and 225 degrees off axis, although this is not a universal rule.
OK, so how do we choose which type of mic to use for our stereo pickup? Well, it depends on what we are recording, plus many acoustic factors in the space where the music is produced.
Frist, what is our goal? Do we want a wide spread of a string section, for example? Or a nice stereo piano sound? Or do we just want to add interest to a sound that is essentially coming from a single voice or instrument?
Define your goals before you start.
Another factor is mono-compatibility. Is that important these days? Certainly not as much as it was back a couple of decades ago when a large percentage of listening was done in mono. Today, most music is consumed in stereo, through headphones or ear buds. But that might not always be the case. Your listener might have a clock radio that is only mono. Or your music might end up playing in a store where it is likely to be mono only.
So, at least some thought should be given to mono-compatibility. You should check your stereo recording in mono frequently to make sure it still sounds good in mono. This will also reveal any reversed polarity sources that could cause problems.
The simplest stereo pickup technique, and one of the oldest, is spaced omni-directional microphones. This was often used in the 1950s for classical music recording and it provides an impressive and wide stereo image that gives a good sense of the recording space.
But it is rarely mono-compatible. Back in the days when this was used for recording orchestras, there was always a single center mono mic that was also recorded. That was used for the mono release of the recording.
Spaced omni mics are also great for adding spaciousness to a mono pickup, or to augment one of the more mono-compatible stereo main pickup techniques.
The distance between the spaced omni mics depends on the width of the sound source, and the dimensions and acoustics of the room. In a large space, they could be 20 or 30 feet apart. In typical studio recording, the mics are usually ten feet or less. Too close together and there isn’t much sense of width. Too far apart and the stereo image tends to become disjointed.
The distance from the sound source also depends on the room and the size of the ensemble. As the only pickup, 10 or 20 feet back would be typical for a group of 30 to 50 musicians or more. Smaller groups would require the mic to be closer.
There are no rules on this, however. You should always depend on your ears to determine the optimum spacing and distance.
This applies to any stereo mic’ing technique.
The technique with the best mono compatibility uses two mics as close together as they can possibly be. This is called a “coincident pair” of mics, and done properly, there should be no major change in the sound when the two mics are combined in mono.
If the mics are in the same space, how do you get any stereo? The key to that is to use directional mics, usually set at a 90-degree angle to each other. Stereo is created because the intensity of the sound reaching the two mics will be different depending on where the instruments are in view of the mics.
This arrangement is often called “X-Y.” It is one of the most widely used stereo pickup techniques because it is simple to set up, gives consistent results, and works well in mono.
Typically, cardioid mics are used for an X-Y pair, but it could be used with mics with a different pickup patterns. Omni mics would not give you much of a stereo image, but sometimes they work. Hyper-cardioid mics usually have a “hole in the middle” effect on the pickup, since there is little sensitivity in that direction.
Nothing says that you have to set the mics at a 90-degree angle to each other. Sometimes a shallower angle works better, especially if you need to narrow the stereo image.
There are variations on the X-Y theme that place the mics somewhat farther apart. Standard X-Y sounds kind of restricted to me, but by spacing the mics a bit, you get a wider image – at the expense of mono compatibility.
On common technique is called ORTF, after the French name for Radio France. This was developed in the 1960s to provide a more interesting stereo spread, while still retaining some mono compatibility. The classic ORTF setup has the two mics angled out at 110 degrees, a bit more than the X-Y 90 degrees. The mics are spaced apart a bit too, by 17cm (6.7 inches).
Personally, I am not a big fan of the sound of X-Y, and I think it is because the off-axis response of the mics adds a weird sound to the pickup. This is because cardioid mics usually have strange pattern differences with frequency, and the off-axis frequency response is usually not even close to flat. But X-Y is always a safe way to record in stereo, with its good mono compatibility and simple setup.
Another popular stereo mic setup is called “M-S.” It was developed in the 1950s as a solution to the simultaneous recording of mono and stereo. It uses two microphones: a cardioid mic facing the performers, and a coincident bi-directional mic at 90 degrees from the cardioid mic. That’s only part of the technique, however. The two microphone outputs must be put through a matrix that separates the outputs to a sum and a difference channels. This converts the cardioid mic (the “Mid” mic), and the bi-directional mic (the “Side” mic) into a smooth stereo image of the performers in front of the mic array.
One nice thing about M-S is that you have quite a bit of control over the stereo image in the mix. You can make the image as wide or narrow as you like.
M-S suffers less from the odd off-axis sound of X-Y, since the bidirectional mic usually has fairly uniform off-axis frequency response, especially if a ribbon mic is used.
Another common stereo mic’ing technique is the Blumlein configuration, named after stereo pioneer Alan Blumlein, an EMI engineer in London who experimented with stereo in the 1930s at EMI’s Abbey Road Studios.
The original Blumlein patent shows a special ribbon microphone with two ribbon elements at 90 degrees from each other and placed in the same magnetic gap. But the configuration that has evolved uses two bi-directional microphones placed in close proximity and oriented 90 degrees to each other. In this practical arrangement, it’s not possible to get the two bi-directional mics very close to each other, so they usually end up a few inches apart. Stereo ribbon mics can get the ribbons closer, but they are still a couple of inches apart or more.
In Blumlein’s design, ribbon mics were used, which are bi-directional in their native form. But the Blumlein technique does not require ribbon mics. Any pair of bi-directional mics will work, and the technique is practical with two multi-pattern condenser mics that offer a bi-directional pattern.
By the way, the bi-directional pattern is sometimes referred to as “figure-8,” due to the shape of its pickup pattern. In the Blumlein configuration, the total pattern resembles a cloverleaf.
For Blumlein, and many of the other co-incident stereo mic techniques, you can use a two-capsule stereo mic. Ideally, the stereo mic should have the ability to adjust the angle between the two mics, and each mic should be multi-pattern. That way you can use the same mic for X-Y, M-S, or Blumlein.
The Blumlein pattern is unique in that it is equally sensitive to sounds coming from in front of the mics and in back of the mics. This might seem to be a disadvantage, and in could be in a room with really poor acoustics. But usually it works OK as long as you are aware of the back pickup.
Note that the stereo image is reversed on the back side.
This back pickup also allows you to arrange the performers on both sides of the mic, which works particularly well for performers who will benefit from looking at each other.
But despite what you might think, this does not provide an omni-directional pickup.
Sure, the system is equally sensitive to sound arriving from any angle around the mics, but the two side pickups have a stereo image that is out-of-phase.
So, it is vital that any sound source that is important to the mix be kept away from the sides.
This out-of-phase side pickup would seem to be a fatal flaw of the Blumlein technique, but in reality, the side pickup of the room reverberation is actually very impressive. It adds a lot of spaciousness to the stereo recording.
However, when combined into mono, the reverberation tends to disappear. Usually that is not a problem, but it should be checked.
A very natural-sounding stereo recording can be done with performers arrayed in front, and in back, of the mics. Just keep everyone away from the sides.
The last stereo recording approach I will discuss today uses an artificial head, sometimes called a dummy head. The concept is to reproduce a close replica of a human head with microphones in the ears. Our perception of sound depends on our head shape, size, and especially the shape of our ears, and the artificial head recording tries to duplicate this. Of course, it is only an approximation. Everybody’s head is different. But the results can be pretty spectacular.
This is also called binaural recording, and it only works well with headphones. You don’t want a “second head” superimposed on the artificial head. So binaural recordings tend to sound pretty strange when played through speakers.
Artificial head devices are made by Neumann and others, although many are designed for industrial testing, not music recording. You can also build one yourself. And there are some DSP software packages that create the binaural effect through sound manipulation of an existing stereo recording.
All of the techniques I have discussed have been aimed at achieving a realistic pickup of an array of performers. But sometimes that is not the goal.
If you want to make a drum set, or an electric guitar, sound bigger and more impressive, you can use a combination of these techniques along with a solid mono pickup (or multiple mono pickups).
Just remember that if you get the mics too far away from each other, you can have phase and time differences, which can smear the sound and make it sound distant.
For an instrument that needs something special to highlight its part, you can use multiple mics set at various distances and angles. Since we have essentially an unlimited number of recording tracks at our fingertips, we can set up as many mics as we want around, say, a guitar amp, and then pick and choose what to put in the mix later.
Often that approach to stereo mic’ing uses various mis-matched mics, which is often an advantage if you want each mic to add a slightly different flavor to the sound.
Even in a classic stereo mic’ing arrangement, there is nothing that requires you to use matching mics. Although this will skew the stereo image in various ways, the result may be more interesting than using matched mics.
Would you ever want to record a single instrument or voice in stereo? Well, I do that all the time. Most sounds do not come from a point. Even a voice has width to it, from the sounds emanating from the mouth, chest, and other parts of the head. A stereo recording of a vocal can put the singer right in the room with the listener.
Stereo is not always what you want, however. Sometimes you need the solid presentation of a mono pickup. It really depends on the song, arrangement, and singer.
How about acoustic instruments? They, too, are not point sources and you can make them much more realistic with a stereo pickup, using any of the described techniques.
There are other variations on stereo mic’ing that I won’t cover here. But these basic approaches will get you started. They are:
Spaced omnis - X-Y - ORTF - M-S - Blumlein - Artificial head - plus any random set of mics
Another aspect of stereo relies on the bleed between various instruments into multiple mics. Although recordists often want each instrument on its own, pristine track, with no hint of any other sound, that is not always the way we end up recording, for better or worse.
Creative use of bleed from many sources into many mics may drive you crazy, but if you know what you want from the start, and monitor your tracks as you would in the final mix, you can capture energy and excitement in the performance that would otherwise be lost.
For example, hearing the drums way off axis from another mic a long distance from the drums can be delightful. This requires a really good-sounding room, however, and a big room at that.
Bleed can also create insurmountable problems, so think carefully about what you are doing and be ready to make changes in the studio, if needed. A room with bad acoustics and a lot of bleed will just make the recording sound really awful and unprofessional. The smaller the room, the worse the problem.
As many of you know, I am partial to Blumlein stereo. It works really well for the type of recording I do, and it fits my style perfectly most of the time.
But I have used all the other techniques at one time or another. Here are my very subjective opinions of each. You should develop your own opinion and preferences, of course.
X-Y recording is very safe and often solves a problem, so I do use it on occasion. But I find it to be kind of “pinched” sounding. It is not a pleasant stereo image to me.
ORTF works well in many circumstances. It’s easy to set up and bulletproof. The out-of-phase characteristic of ORTF at the higher frequencies bothers me. It lacks the solidly of Blumlein or X-Y.
I listen to what others do with M-S and it sounds great to me. But whenever I have tried it, I have concluded that it just doesn’t fit my recording aesthetic. I’ll keep trying it, however. whenever I have the time to experiment in a session.
The artificial head approach is quite specialized and it is ineffective for a general audiences, although with most people listening on headphones or ear buds these days, it might have more application than it did in the past. Personally, I am unimpressed by that sound.
I like the effects you can get with random mics in stereo. It can add interest and drama to an instrument. I use that approach often. It’s tricky to find the right combination of mics, distances, angles, etc. and I usually do not have time to do as much experimenting as I would like.
Often, we may set up two or more mics for an instrument or voice, so we can experiment and decide what works best. This offers us an interesting opportunity. Once the best mic is selected, leave the others in the vicinity and listen to what they do when added to the mix. An off-axis and perhaps distant pickup with a mic with different characteristics can often be blended in to make the instrument, or vocal, sound really different and interesting.
On some instruments, I like to use two very different mics placed close together. The mics might be cardioid condenser with a lot of exaggerated high-end, plus a bi-directional ribbon mic. I use the two mics panned hard left and right, and the result can be spectacular. The mono compatibility is usually very good, too, if the mics are close enough to each other. It can widen the apparent image of the instrument, or vocal, and yet retain a solid sound. Of course, you can always choose one or the other to use alone, of course. Sometimes I use one of the mics to send to the reverb only and use none of the direct sound.
I find that sometimes the most interesting sounds come about through a random mic situation that was not planned, but just turned out to work great.
That keeps recording interesting.
And sometimes a well-isolated, mono track is still the best choice.
Thanks to listener Bill Sallak for suggesting this topic.
You can suggest topics, too. Just drop me a note: firstname.lastname@example.org
Thanks for listening, subscribing, and for all your comments.
This is My Take On Music Recording. I’m Doug Fearn. See you next time.