Cogging!

The facts are that almost all motors/drives cog to some extent.  Its caused by the finite number of torque pulses delivered to the platter.  Its similar to the smoothness of a car with a 4 cylinder motor vs a car with a 6, 8 12 or 16 cylinder engine or even better an electric motor.  Generally the more cylinders/poles, the smaller the torque pulses and the smaller the resulting variation in speed of the platter/car that are felt as vibration or lack of vibration or smoothness.  There are lots of ways to reduce it given a number of poles/cylinders.  Some 4 cylinder engines (BMW) are a lot smoother than others.  So if the cogging is below an audible level, its not a problem.  Usually it shows up on piano music.  Wow and Flutter are also forms of speed variation that are specified.  Typically they are lower in frequency than the cogging caused by a direct drive.  In a TT with a belt, the belt and the inertia of the platter act as a mechanical filter to reduce it.  The electronic drive to a direct drive TT can be designed to reduce the cogging also--as the above poster mentioned, reducing the voltage after startup is a simple way to do this for AC synchronous motors as well as steppers.  An oil bearing support versus low friction ball bearings is another thing that can help.  As with all things, its a balance and most generalizations have their qualifications.  Execution and fine grain detail design is required.  

And a coreless motor helps most of all.

Its a new dance try it.

The facts are that almost all motors/drives cog to some extent. Its caused by the finite number of torque pulses delivered to the platter. Its similar to the smoothness of a car with a 4 cylinder motor vs a car with a 6, 8 12 or 16 cylinder engine or even better an electric motor. Generally the more cylinders/poles, the smaller the torque pulses and the smaller the resulting variation in speed of the platter/car that are felt as vibration or lack of vibration or smoothness.

This is not accurate.  Cogging is caused by the change in variable reluctance as the rotating permanent magnet rotor passes over the metal pole pieces and air gaps between them.  The drive signal is not "pulsing", it is a continuously smooth sine wave and the rotor synchronizes to the rotating field;  if there were no metal pole pieces in the stator (coreless motor), there would be no cogging.

Adding more poles to a motor does not reduce cogging, the attraction between the rotor and the stator poles remains the same.  It may feel "finer" in a 300 RPM motor vs 600 RPM, because there are twice as many poles, but the motor spins at half the speed so the vibrations will be identical if the motors are the same power rating.  This was measured and discussed here:
https://www.diyaudio.com/forums/analogue-source/309925-hurst-motors-300-rpm-vs-600-rpm-upgrade-myth.html

The metal poles of the stator concentrate the magnetic field developed by the stator coils.  The higher the drive signal, the higher the magnetic field and the stronger the attraction to the rotor and the higher the cogging.  The vibration in an AC synch motor is directly proportional to the power consumption, not the number of poles.  Reducing the drive voltage will reduce vibration, but it also reduces torque, this is only a band-aid and can start to have a negative effect on the sound.

Skewing the angle of the stator gaps can greatly reduce cogging as the rotor is never completely aligned with the poles or the gaps.  Most 3 phase BLDC motors are constructed this way.  When driven as 3 phase AC synch motors, they exhibit little or no cogging and are as smooth as coreless motors.