[daip] DTSUM and DiFX data

[Eric Greisen]

Walter Brisken wrote:
> DiFX, the new software correlator that is to replace the VLBA correlator 
> in the next few months, can produce data with integration times that are 
> not restricted to be multiples of 131072 microseconds.
> 
> DTSUM seems to assume that it is an exact multiple of this:
> 
> Data integration time (cross-power) =  2.000000 =   15 * 0.131072 seconds
> Data integration time (total-power) =  2.000000 =   15 * 0.131072 seconds
> 
> I'm not sure if displaying the "=   15 * 0.131072" is for information only 
> or if internally it (or any other VLBI AIPS task) makes any assumption 
> about the granularity of integration times in actual calculations.  I 
> hope not!
>

The 131072 number occurs 2 places in AIPS - this display function in 
DTSUM which I think was part of the vlbi simulator set that Athol began 
but never finished (DTSIM, DTCHK).  It also occurs in KRING as the 
default for CPARM(1) a global integration time (16 * 0.131072).  It has 
no specific use in other words.

Eric Greisen