The telescope software logs the results of pointing measurements upon explicit
command, but what is recorded are the extant values of the relevant parameters,
rather than those applicable at the time of, say, mid-event, which is perhaps
the more relevant instant. In particular, (az,el) may be quite different.
Recent analysis of
tracking data has compensated for the delay by interpolation, but, to date,
regular pointing data has not been corrected. The correction in time
amounts to about half the duration of a FIVEPOINT experiment, i.e.
about 1
, in which time a source at 75
elevation moves through
1
during transit. This certainly equals the resolution of the track model,
and prevents track model analysis on smaller scales in these azimuth regions.
However, again, only in this zone of avoidance is this problem particularly bad.