Stereo-imaging tools are often included in DIY mastering packages or equalisers with stereo spreading facilities (most commonly offering some kind of mid-side processing options), so it is important to acknowledge their purpose and their limits.
As with many of the specialist tools we have for processing audio, they are great at solving specific problems. If you have something mono or largely in mono, for example, and you need to try and widen it, you can add reverb or perhaps exaggerate the little stereo information that already exists in the track.
But what exactly is the ‘stereo information’?
Well, it has to do with the relationship between the ‘out-of-phase’ information, and the ‘in-phase’ information. Anything that is in-phase happens at exactly the same time in both channels, and that information will appear to be centered. If anything is slightly delayed off to one side or the other, compared to the center of the image, it is ‘out-of-phase’ and is one of the things that creates a stereo sense of spread.
So when you are mixing, you are using pan pots, delays and reverbs (etc) to create a stereo image of individual elements within an overall stereo mix. However, when you go in during mastering and increase the out-of-phase component, you are changing the relationship between the out-of-phase and in-phase parts of the signal of the entire mix. You therefore are able to radically change the sense of the stereo image and the placement of each individual instrument in a song, which can be very dangerous if not treated with care. And although you may increase the perceived wideness, this is at the expense of the in-phase components in the song. That is to say, the elements that are dead-center, which also tend to be the most important elements of most productions – vocals, snare, kick, bass – are weakened.
So sometimes – in its various forms – stereo-imaging can be used to good effect. However, one should err on the side of caution.