Dither question

[spokes...@gmail.com]

For many years, I assumed that dithering maintained the original guide star position as a reference and randomly selected a new position some distance from that initial frame of reference. I thought I had read somewhere that this was how dithering worked and was why after a mount move, PHD2 guides to the intended dither position curing the settling process. If it always knew where the dither landed (within the user-specified settling tolerances) it would always know roughly where it was with respect to the starting frame of reference and the dithering would take place within the vicinity of that frame of reference.

The alternative would be a true random walk where each dither is performed with respect to the last dither only, and any relationship to the original frame of reference is lost after the first dither. This would result in an overall random drift equal to the average dither distance times the square root of the number of dithers. I don't *think* that I have noticed that much overall drift, though most imaging sessions are interrupted by a meridian flip and plate solve which would reset everything to roughly the same initial frame of reference when the session began. So maybe drift from a true random walk would not be as obvious as I thought it would be. 

So, which is it - does PHD2 do pseudo-random moves that intentionally remain in the vicinity of the starting frame of reference, or it is a true random walk that will drift as the square root of the number of dithers over the course of a session. If the latter, why bother guiding to the intended dither destination during settling? Why not let the mount lant where it lands?

Thanks.

[Bruce Waddington]

For random-mode dithering, each random move is relative to the current lock-point which, in turn, is the position of the previous dither.  I can't parse the last part of your question.  A dither is defined as setting a new lock-point and guiding to that position.  

Bruce