Progress towardhigher digital areal densities is related to the data track position sensing accuracy obtainable in future disk drives. In most systems available, position sensing is done remotely from the data track and subject to different dimensional stability conditions. To obtain higher track densities, position sensing reference information must be moved near the data track. Several techniques are available to embed reference information in the data track. All the techniques described have one of two basic principles of operation. One method uses comparison of reference signal amplitudes and the second method uses arrival time differences of reference signals. The time method is relatively unknown but offers the designer some new alternatives. This paper describes an overview of several methods. Relationships are derived to relate the influence of noise on both basic methods. Linearity, relative hardware simplicity, and track capture range are also discussed. It is our conclusion that the Amplitude or Tri-Bit method offers the greatest hardware simplicity, but is limited by noise and non-linearity at high track densities. The time dependent or Chevron method requires two data channels, but offers a relatively higher noise immunity, better linearity, and higher data capacity.