[daip] True Stokes

[Eric W Greisen]

Mark Reid wrote:
>
> Eric,
>
>    There is somethings that I don't understand about the "true Stokes" 
> conversion when writing a FITS tape.
>
> 1) FITTP has "DOSTOKES", but FITAB doesn't.  Does FITAB also convert 
> to true Stokes?
>
> 2) I don't think I have ever set DOSTOKES>0 (it defaults to -1), and 
> yet I seem to have tapes with Stokes I,Q for data that started RR,LL.
>
> 3) If one converts RR,LL to I, one must apply calibration tables 
> before converting.   Has that been done?   Assuming it has, I don't 
> see how the CL tables that are on the tape of any value (and actually 
> could cause trouble)?
I forgot to get back to you,  I note that most tasks that allow calibration
will convert the Stokes.  FUDGE is the most simple minded of these but 
SPLIT
and others will also do this for you.  The problem is that the weights 
will no
longer be what they should.  Conversions keep assuming that adding 2 numbers
(divided by 2) makes them more reliable - so the weights go up with each 
conversion.
The RR/LL -> I nay even be correct in this but one can't go back.

Anyway - no new task is needed.

Sorry about forgetting to reply.

Eric