[daip] OMFIT problem

[Eric Greisen]

On 06/21/2017 11:37 AM, mz wrote:
> Dear Eric,
>
> I did ask you about AIPS problems years ago. Now I still find it is
> difficult to google out all the answers about AIPS through your daip
> mail list. So I decide to write you an e-mail directly again.
>
> As a radio astronomer and a loyal user of AIPS, I intend to do all my
> data reductions within its capability. I want to do some astrometry
> measurements on an HST image with AIPS. It is understandable that the
> FITS file was piped out from IRAF which is not necessarily compatible
> with AIPS. I have hacked the keyvalue of "CDELT1/2" in the FITS header
> to get the correct "coordinate increments".  However, when I was trying
> to use TVSTAT to measure the flux in the certain region it reads out an
> unit of "COUNTS/S". I tried to grep through the AIPS source code to find
> out where the BUNITS becomes "COUNTS/S" but the AIPS source code looked
> totally mystical to me. I could only find where the BUNITS becomes
> "Jy/beam".
>
> So may I ask you for a favour to tell me what does the unit "COUNTS/S"
> mean here. My bold guess is "counts per second" but it doesn't make any
> sense to me right now. Cheers!
>
> Regards,
> Ming

I suspect that the counts per second unit was provided by IRAF and it 
would be similar to Jy/pixel except that there is an unknown calibration 
factor.  Thus you cannot get a total flux in units that one can 
understand.  One has to know how to convert counts/s to an intensity and 
that depends very much on the instrument, wavelength, etc.

Some images from IRAF have good spatial coordinates and do not have to 
be hacked.  They do however sometimes contain a skew in the coordinates 
which AIPS does not handle generally.  The AIPS task DSKEW was written 
to handle this.  However, your images may not have had useful 
coordinates since some optical, IR, and x-ray instruments do not produce 
them in all cases.

Eric Greisen