[daip] car coordinates

[Eric Greisen]

Lawrence Rudnick wrote:
> Eric - I think I wrote to you about this before, but can't find the old 
> mails.  We are using DRAO all sky Q and U images, which are in cartesian 
> coordinates that AIPS does not recognize. [ 
> http://www.drao-ofr.hia-iha.nrc-cnrc.gc.ca/26msurvey/data.html]
> Maik Wolleben, who made the original images, is trying to help us, but 
> doesn't know how to make "all sky" images in sin or tan.
> # We are looking for the very simplest thing we can have him do that 
> would allow us to use his images in AIPS, and make use of OGEOM, FLATN, 
> etc. as needed to create whatever we want.  What is the simplest thing 
> we can ask him to do?  Perhaps breaking up his images into small 
> chunks?  Perhaps going into AITOFF which I think he can do, and perhaps 
> we can get back out of to compare individual regions (I'm a little 
> worried about losing information at high latitudes because of the 
> odd-shaped pixels - not clear we could recover).
> 
> So, your advice welcome, we just need to make it easy for him since he'd 
> be doing us a favor.  Many thanks.  lr
> 
I do not know what to recommend since I do not know how one can do 
all-sky in -CAR - at least the poles would get very messy.  One cannot 
make -sin and -tan in all sky.  I guess that AIPS would see a -CAR with 
ref RA 0 and ref Dec 0 as a linear coordinate and so if you drop the 
---CAR and --CAR from the header types AIPS might be able to handle it.
Any other ref value suggests to me that DRAO is misusing the 
coordinates.  One thing to watch - I have had many people claim that 
their image is cartesian and I find out it is what aips calls -GLS
where cos(dec) enters in the x axis.

I am cc Mark Calabretta since he understands all this better than me.

Eric Greisen