mp3 based motor drive?

If you have an audio generator you can do it that way too. The real issue is that the generated frequency has to be very stable! The best way I have heard of is to use a high frequency crystal-controlled oscillator where the crystal is heated in a small oven (a once-common practice in radio gear) and then the resulting output is divided down to 60 Hz. This can result in the frequency having stability to several significant digits.

Atmasphere is correct, but not with every type motor. With those that are frequency controlled, however, a good clock is a necessity. A typical quartz crystal is capable of accuracy of around one part per million, and an OCXO like Atmasphere mentioned will arguably take the accuracy up a lot of notches, and keep it almost perfectly stable. Now, if you really want to get serious, add GPS referencing to the OCXO. Then, we are looking at an accuracy of around one part per trillion, and unparalleled dead-on stability. The obvious question to that approach is whether anyone can detect the difference. If you consider that inherent tracking error of the idler wheel rubber in a rim drive turntable is around one part per million and cannot be mechanically improved, the commonplace quartz clock may be enough for very tight control. Belt drives have creep, so there's another obstacle to hurdle.

Then again, audio guys like to push the limits. A GPS heated clock isn't really that far-fetched when you take into account what we do in this crazy hobby.

Can an Ipod be used to do it? The short answer is probably, but you'll need to add a bunch of stuff with it.