Brief Summary

Stereo recording captures sound from two or more sources and creates a more immersive and natural soundfield than mono recording. In this article, we'll go over the necessary steps to record audio in stereo, including mic choice and arrangement and information on Stereo Recording Angle (SRA) and its effect on the stereo panorama.

Going Deeper

Should I Record In Stereo?

Stereo sounds are more impressive than mono sounds so it would be easy to think that stereo is ‘better’. Stereo reproduction is usually better than mono, but in a multitrack project it doesn’t follow that every element needs to be in stereo. For example vocals aren’t tracked using stereo techniques and they are the most important element of most projects whereas a piano will almost always be recorded in stereo. Why is that?

The difference is one of width. A singers mouth is effectively a point source. The only stereo information you can capture beyond a solo singer’s left/right position when recording it is the contribution of the room in which the recording is being made. For most contemporary music this is undesirable and a dry recording with artificial reverb is far more common, however for a classical recording this wouldn’t be the case.

This differs from a piano in that a piano has appreciable physical width and a stereo recording is more appropriate. That width can be captured using any of the many stereo miking techniques available. But which technique should you choose and why?

Once you have established that the source you are recording is a suitable candidate for stereo recording, here are the steps you’ll need to follow:

1. Choose your microphones - These should be a pair of the same model. While a matched pair of microphones, meaning a pair of mics which have been measured and identified as being within a specified tolerance of each other is of course a good thing, in reality a pair of the same model of mic should be OK. Modern manufacturing means that most good quality mics are well matched. As long as gain is properly matched and the mics sound the same to you then you’re probably fine.

2. To record in stereo you'll need at least two microphones. Stereo mics which combine two mics into one unit are available and are very convenient but using two separate microphones offers greater flexibility. Condenser mics are the default choice but ribbons can also be excellent. Multipattern mics offer the greatest flexibility but while fixed pattern mics limit choices, a stereo recording can be made with any pair of identical mics. If your mics don’t match you can still make a stereo recording but the results will be compromised, probably seriously.

3. Choose your technique: XY, Spaced, Mid/Side, Near-Coincident, Binaural etc.

4. Connect your microphones to your interface, you’ll need a pair of inputs with mic preamps which have adequate gain and low noise (which might be an issue if the thing you are recording is quiet and/or distant) and if you are using condenser mics you’ll need 48v phantom power on the mic preamp to power the mics. You might also be able to link the channels for stereo operation.

5. Position your microphones. Choosing the right position and auditioning the results is a crucial step so allow time for this. There is a lot which can be done post-record to salvage mistakes but it’s always better if it’s done right in the first place. There are a few things to consider when placing your mics so we’ll return to this point.

6. Once your microphones are positioned correctly, you'll need to adjust the input levels on your recording device. Try to make sure the levels are equal for both microphones, and that you're not clipping (i.e., the levels are not too high). To get an approximate setting for gain I’ll usually set one mic’s level appropriately, leaving adequate headroom to avoid clipping, I’ll typically leave as much as 24dB of headroom when recording at 24bits. This is conservative but I’d recommend leaving at least 16dB between the peaks when auditioning as players tend to get louder. Once this is done match the level with the second mic by getting the player to clap or hit a drum (if they are a drummer) and match the level of the first mic with the second using a peak level reading from the track’s meter. There are more sophisticated and accurate ways to set gains but this method is quick and simple.

7. Record!

Which Technique Should I Use?

There are lots of different ways to set mics for capturing stereo but broadly they fall into two categories - coincident and spaced. Stereo information can be based on level differences between the two speakers, this is how a pan pot works, or timing differences. The ear which hears a sound first indicates which direction the sound is arriving from. This precedence effect can be captured by placing mics at different points in space and level differences can be captured by using directional mics pointing in different directions.

Coincident - Level-Based Stereo

XY is the best known level-based stereo technique. This typically uses two cardioid mics placed as close as possible to each other but at 90 degrees to each other. Other patterns and mutual angles can be used but they are all ‘coincident’ techniques in that the two mics coincide in space (i.e. are in the same place). Other coincident techniques include Blumlein, which uses fig 8 mics and Mid Side which uses a forward facing cardioid and a sideways facing fig 8 along with some clever routing to give stereo.

Spaced Pairs - Time Of Arrival Differences

Placing two mics at different points in space captures different times of arrival, as a sound source will have different distances to cover to each mic. The simplest example is a pair of omnidirectional mics, which capture only timing differences. However directional mics can and are used in spaced pairs which adds an extra element to the nature of the captured sound. More than two spaced mics can be used in an array to capture large ensembles but this falls outside the scope of this article.

Near Coincident - Mixes Both Approaches

Near coincident techniques are popular as they offer some of the benefits of both approaches. Popular variations include NAB and ORTF, named after the organisations which standardised them. There is nothing magical about these arrangements, they are just variations which work well and have been standardised so that they can be accurately repeated.

Binaural - Mimics What The Ears Hear

Lastly I’ll mention binaural. The cues the ear hears to infer spatial cues are complex, involving timing, timbral differences caused by the influence of ear shape, masking caused by the acoustic shadow of the head, and phase effects. A facsimile of this can be imposed on a recording using an impulse response called a Head Related Transfer Function (HRTF) but the simplest way to capture these cues when recording real sounds in real spaces is using a ‘dummy head’ mic which, as the name implies uses an anatomically accurate head, complete with ears to house microphones. The AMSR videos which are so popular online have introduced a whole new group of people to the startlingly accurate nature of binaural recordings. They only work when played back on headphones though, though with the rise of portable devices this is less of an issue that it was in years gone by.

Stereo Recording Angle (SRA)

One important consideration when positioning microphones for stereo recording is the Stereo Recording Angle (SRA). The SRA describes the relationship between where sources are relative to the mic array and where those sources will be placed between the speaker on playback.

This SRA is affected by the mutual angle between directional microphones and the spacing between spaced mics. If the SRA is inappropriately wide it can cause sounds to crowd around the position of the speakers, leaving little in the centre, too small an SRA can result in the stereo image not filling the entire panorama between the speakers. If you want to set your array correctly there are excellent tools available which can help, for example the free Recording tools App from Neumann or Image Assistant from Schoeps.

Recording is stereo isn’t difficult but there are a couple of common mistakes people make if they are trying to record proper stereo, by which I mean capturing an accurate representation of a real sound in a real room. You can of course create arbitrary amounts of width using multiple mics, there is nothing wrong with this kind of ‘pseudo stereo’ which you might make by placing two different mics at different points around an acoustic guitar and panning them left and right to create a flattering sound. However if you are trying to capture stereo realistically make sure your mics are the same, your gains are matched and your mic placement is appropriate to capture the full width of the sound source(s) so that your stereo array will produce an appropriate placement of these sounds across the stereo panorama.

Stereo can seem complicated if you’re trying to do it properly but its immensely rewarding when you get it right!