The Stereo Image

The brain determines stereo spacing via a number of different methods. The 2 most prominent are time differences and amplitude differences. As our ears are placed on opposite sides of the head. a signal to one side will reach one ear before the other - our brain automatically decodes this so that we know the sound is nearer one side than the other. Similarly, a sound that originates on our right hand side will appear, depending on distance, louder to the right ear than the left ear - again our brain can interpret this to allow us to make an assumption on the position of the sound source.

When working with a mono sound source, for instance an electric guitar track that was captured using one microphone next to the amp, the pan control allows you to position the guitar anywhere on the 'horizontal' axis. When the pan control is turned completely to the left, then the signal will only be reproduced by the left loudspeaker - correspondingly the pan control set fully right will ensure the signal is only reproduced by the loudspeaker on the right. Unsurprisingly, using this logic, setting the pan control to the Centre will mean that the signal is reproduced in both speakers with equal amplitude.

What should be pointed out here is the presence of what we call 'Panning Laws'. Traditionally, on analogue desks, when a signal was panned central it would appear louder than if it was panned hard left or right - this was due to the linear gain changes between channels when using the pan control. Most modern software applications let you counter this by setting a 'Panning Law' that suits your method of working. Possibly the most common setting is -3dB - this ensures that a signal will appear to be the same amplitude regardless of where it is panned.

Back to our mono guitar signal - without further processing/recording a mono signal will not be able to give the sense of being 'stereo'. Often this can be counteracted by double-tracking the part (i.e. recording the part twice) and then panning the 2 mono signals left and right - the subtle differences in the performance of the musician will ensure the ear differentiates between the 2 signals and the use of panning will give a sense of space to the performance.

If you do not have the luxury of double tracking there are many ways a pseudo-stereo image can be created. One of the most common methods is to make a copy of the track and then delay it by a few milliseconds and pan each signal left and right. This method is somewhat haphazard however due to problems with phase. An alternative approach is to make a copy of the signal and treat one of the signals with subtle chorus, flange and/or phaser. This creates a somewhat random relationship in the time domain between the wet and dry signal and is in most cases preferable to the previous method. It is also possible to create a faux stereo effect by making a copy of the audio signal and EQ-ing each signal differently.

When recording sound sources such as acoustic pianos and acoustic guitars it often makes sense to actually record in stereo giving us 2 audio files to work with. Obviously to record in stereo we need to use 2 microphones to pick up the sound.

When recording in stereo there are a number of stereo recording techniques that are commonly used by engineers, although the 2 most common for the home engineer are 'coincident' and 'spaced' pairs. These 2 methods work in different domains, amplitude and time.

A coincident pair, sometimes called an 'XY Pair' or 'Crossed Pair', are placed so that the microphones capsules are as close to each other as possible - usually on top of each other - but are positioned at right angles to each other (see image on right) To achieve this positioning it is often easier to use smaller microphones without shockmount's but obviously this is determined by the equipment on hand and the sound source being recorded. Although it varies, this approach is generally best suited to using directional polar patterns such as cardioid, hyper cardioid or similar This close placement of the capsules ensures that the sound reaches both microphones at the same time (or as near as possible) As there is very little difference in the time domain, the main difference between the 2 signals is the different levels in terms of amplitude. The signals are subsequently panned left and right (use your ears to find the best placement) and should give an adequate sense of space. One obvious disadvantage of this method is that the lack of differences in the time domain makes the stereo image less 'obvious'. However this is somewhat negated by the fact that mono compatibility is greatly increased and less problems are likely to be caused by time/phase issues which can result in unpleasant comb filtering and/or phase cancellation when the signal is reproduced on mono sound systems such as portable radios or mobile phones.

The alternative method is to used a 'spaced pair', as seen in the diagram on the right. In this configuration the microphones are placed a distance apart (the further the distance, the more exaggerated the stereo spread) pointing at the sound source - with this setup an omni-directional pattern is often favored but other directional patterns may be used. As the capsules are spaced, there are more time discrepancies between the signal being recorded by each microphone in a similar way to the way that our ears pick up sound at slightly different times. This produces a more 'obvious' stereo spread when the signals are panned left and right on the console/DAW. However, the downside of this approach is that mono compatibility can be reduced and phase issues, which can manifest as a lack of focus or clarity in the outer reaches of the soundstage, are more prevalent.