[daip] Change in cross-correlation visibility weights in (VLBA) UV files?

[Daniel Lebach]

Hello,

For the past few years, the maximum visibility weight I have found in
any UV file of VLBI data correlated at Socorro with 2.883584 s
integration time is 1.57142857 , or 22/14, or more generally,
(integration time)/(0.131072 * 14) .  The most recent experimental
data I have processed, however, which is for BR080C and was
correlated (I think) on 5 Dec. 2002, has an integration time of
3.014656 s (23*0.131072) and a maximum visibility weight of about
3.2857 .  Clearly this weight does not obey the formula I have relied
upon for the last few years.  Has the formula on which visibility
weights are based changed?  If so, can you tell me the NEW formula?

These weights are historically correlated directly with the number of
correlated bits, and thus I have used them in data SNR calculations.

Regards,
Dan