[daip] X-Y mounts in FITS-IDI (and maybe AIPS too)

[Walter Brisken]

Hi Eric,

Thanks for the clarifications.

It is now clear that XY in FITS-IDI implies XYEW, and I do now have some 
recollection that Richard Dodson did some work to get that into AIPS, 
perhaps a while back.

I know of no still-operating XY mounted antenna that is not EW, so there 
is no urgency in putting in support for XYNS.  It would be good if this 
could be better documented somewhere.  I'm happy to spew massive warnings 
in difx code that warn that XYNS is not supported downstream.

If you would like to put in the support for XYNS I'd be happy to make sure 
DiFX can properly populate the tables.

Let me know your preference...

Is parallactic angle calculation the only thing AIPS needs mount type for?

-W

On Tue, 24 Aug 2010, Eric Greisen wrote:

> Walter Brisken wrote:
>>  Hi DAIP,
>>
>>  Adam Deller has been digging up quite some issues with X-Y mounted VLBI
>>  antennas within the correlator.  We have noticed something in FITS-IDI
>>  that likely couples to AIPS as well.  In AIPS memo 114, the reviced
>>  FITS-IDI specification, the mount type can be Alt-Az, Equatorial,
>>  Orbiting, Nasmyth (R and L variants) and X-Y.  Unfortunatly X-Y mounts
>>  come in a continuum of flavors that depend on the orientation of the lower
>>  axis, with the East-West and North-South being the only actual variants I
>>  know about.  Since FITS-IDI cannot convey whether the X-Y mount is E-W or
>>  N-S (or something else) it has no hope of getting parallactic angle
>>  calculations correct in general.
>>
>>  A proposed solution might be to add two more mount types (XYNS and XYEW)
>>  to the FITS-IDI list with new identifying numbers (possibly 7 and 8),
>>  leaving the ambiguous XY as currently defined at a value of 4.
>>
>>  To be useful, AIPS will have to make use of this information.  Do any of
>>  you know which variant of X-Y is currently assumed in the parallactic
>>  angle calculations?
>>
>>  -Walter
>>
>>  _______________________________________________
>>  Daip mailing list
>>  Daip at listmgr.cv.nrao.edu
>>  http://listmgr.cv.nrao.edu/mailman/listinfo/daip
>
> The code for PARANG is below
>
> C                                       Parallactic angle
>             PANGLE(LOOPAN) = ATAN2 (COS (ARLAT) *
>      *         SIN (HRANG), (SIN (ARLAT) * COS (DECAPP) -
>      *         COS (ARLAT) * SIN (DECAPP) * COS(HRANG)))
> C                                       Compute only for X-Y mount
> C                                       (Hobart only?)
> C                                       Co-parallacti
>             IF (MNTYP(LOOPAN).EQ.3) THEN
>                if (first) then
>                   WRITE (MSGTXT,1000)
>      *               'Using CO-PARALACTIC ANGLES for ', LOOPAN
>                   CALL MSGWRT (2)
>                   FIRST = .FALSE.
>                   END IF
>                PANGLE(LOOPAN) =  ATAN2 (-COS(HRANG),
>     *            -SIN(HRANG) * SIN(DECAPP))
> C                                       Nasmyth
>             ELSE IF ((MNTYP(LOOPAN).EQ.4) .OR. (MNTYP(LOOPAN).EQ.5))
>      *         THEN
>                DPARM = TIME
>                CALL SOUELV (LOOPAN, DPARM, HA, EL, AZ)
> C                                       Right handed is FA=PA+EL
> C                                       Left handed is FA=PA-EL
>                IF (MNTYP(LOOPAN).EQ.5) EL = -EL
>                PANGLE(LOOPAN) = PANGLE(LOOPAN) + EL
>                IF (FIRST) THEN
>                   WRITE (MSGTXT,1000) 'Using Nasmyth formula for ',
>      *               LOOPAN
>                   CALL MSGWRT (2)
>                   FIRST = .FALSE.
>                   END IF
>
>