RE: FITS header values

["Albertson, Chris" <CAlbertson@primeadvantage.com>]

I was about to comment that I was surprized the ra and
dec image scale were the same as I thought Tom was looking
up above the equator then I read Arne saw it first.  
I think you need to multiply by the Cosine of the declination.

With our wide field, the image scale will be nonconstant
and in "ra,dec space" the field of view is not square
but the CRDELTn value should apply at the reference
pixel (CRPIX1, CRPIX2)

The reference pixel need not be an integer either.  If can
be anyplace.



> -----Original Message-----
> From: aah@nofs.navy.mil [mailto:aah@nofs.navy.mil]
> Sent: Tuesday, April 10, 2001 8:12 AM
> To: tass@listserv.wwa.com
> Subject: Re: FITS header values
> 
> 
>   I have been working on software for the IR camera
> and have not had time to participate in this
> discussion.  It is snowing now, so I have a few moments
> to write emails and continue with the GRB010222 paper.
>  One problem with Michael's definition of CDELT2
> (the RA increment).  This would then change at every
> new declination setting.
>   Before you say 'fooey', you should read the new
> standard for WCS thoroughly and make sure you
> understand it, as there are usually reasons why
> things change.  Since the Mark IV will run for the
> next 5-10 years (it will at NOFS, anyway), I think
> we should do it right where possible.
>   While keyword order makes no difference as far
> as most FITS readers are concerned, it does make
> the header cleaner and easier for a human to read
> if you group like parameters together.  For example,
> Tom is including  DATE and DATE-OBS; place them on
> adjacent lines.  Same with UT, UTSTART and TIMESYS
> (I normally consider them in the 'date' group);
> coordinates; temperatures; etc.
> The comments about FITS Definition documents should
> be placed at the end of the header.
>   BZERO and BSCALE should definitely be in the fits
> header.  I don't recall; perhaps cfitsio adds them
> automatically when it writes the file.
>   The easiest and most practical way to get proper
> UT is to run your computer on UT.  You never have
> to worry about conversions or daylight savings time.
> Arne
>