The turntable's job is to spin the LP at a constant speed while isolating the LP-arm-cartridge from the room. It is a deceptively complex job.

For whatever reason, Direct Drive turntables rarely have decent isolation. There are acoustic feedback and gross mechanical feed through problems. There is no fundamental reason why this must be true. Perhaps the Direct Drive designers don't think isolation is important or can't accomplish the objective.

Poor isolation has a profound impact on the sonic outcome. We remember one rather crude turntable, a belt drive unit, where one could drum fingernails on a wooden floor about five feet away from the turntable (mounted in a rack) and clearly hear the result through the speakers.

The lure of a Direct Drive motor is that it runs at 33.33 RPM or about 0.5 revolutions per second. As motors turn, since they are imperfect in many ways, they will emit noises with each revolution. If the motor spins fast (600 or 1200 RPM are common speeds -- yielding noise at multiples of 10 or 20 Hz) the noises will fall into the audible range or a range that may excite arm-cartridge-suspension resonances. At about 0.5 revolutions per second one would think that any noises developed by a direct drive motor would be insignificant.

Unfortunately, there is no easy way to design a motor that runs at 33.33 RPM. One must use a complex motor where the internal magnetic field races faster than 33.33 RPM while the mechanical motion occurs at 33.33 RPM. If you carefully examine how the torque is applied to a direct drive motor, you'll find it arrives in small packets -- similar to a potter's wheel that is kicked from time to time while a large flywheel keeps the motion relatively smooth. In our turntable, these pulses of energy (or a harmonic) may excite the arm-cartridge-suspension resonance causing audible problems. In some of the low end, fairly crude Direct Drive turntables, you can easily observe cyclic speed changes within each revolution. The designer ignores these speed changes because the problems they cause are believed to be smaller than the speed changes caused by warps and eccentricities common to most LPs.

The belt drive designs try to isolate the motor from the arm-cartridge-suspension system by coupling the motor to the platter through a limp rubber band. In theory the belt will not transmit much motor noise energy to the platter.

Unfortunately, many of the motors are so cheap and poorly made, that the belt is overwhelmed and transmits enough noise energy to cause trouble anyway. There may also be significant motor noise energy transmitted to the arm through the turntable suspension. In many cases, even on big name high end turntables, we can slip off the belt, park the stylus on the stationary LP and clearly hear the motor running. Some other motors have an external magnetic field that induces hum directly into the cartridge.

Bottom line: We can't make a blanket statement; there are good, bad, and ugly examples of both Direct and Belt drive turntables. Unfortunately, there are very few really good turntables.