Stereo imaging. How is it done?

I guess this question is for a recording or mastering engineer, do we have one in the house? How is the center image accomplished? Is the same technique applied to place instruments to the left and right of the center image between the speakers? Is this done during the actual recording of the material or is it done later during the final mastering? It blows my mind when its done properly but on some recordings I have to compensate with the balance and move the center image over a foot or so. My only guess as to why this happens is that my speakers use first-order crossovers and the sweet spot is very narrow and perhaps different studios use monitors that are designed for multiple listners therefore making it harder to define the images. Or mabye they are just bad recordings.

Also thought I might get a few pointers for a friend who has a studio at home.
There are two ways to make a stereo image. The simplest is to use two microphones (or a single stereo mic), place them appropriately and record. This technique will capture whatever L/R and depth information present in the recording space. The more commonly employed method is to independently record multiple mono tracks and then in the mixing process create a synthetic stereo image by altering L/R channel volume and phase/timing. When done with care, the synthetic stereo can be very convincing.

A center image is equivalent to a mono signal with equal volume in each channel generated by each speaker at the same time. Increasing the volume in one only will pull the image towards that speaker. Also delaying the signal will pull the image towards the opposite speaker. Artificial image depth can be created by various reverb/modulation programs.
With two channels: pretty poorly.