[daip] [!9033]: AIPS - CONVL - issues with convolving beam and Jy/beam units

[Michael Bietenholz]

Michael Bietenholz updated #9033
--------------------------------

CONVL - issues with convolving beam and Jy/beam units
-----------------------------------------------------

           Ticket ID: 9033
                 URL: https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/9033
           Full Name: Michael Bietenholz
               Email: mbieten at yorku.ca
             Creator: User
          Department: AIPS Data Reduction
       Staff (Owner): -- Unassigned --
                Type: Issue
              Status: Open
            Priority: Default
                 SLA: NRAO E2E
      Template Group: Default
             Created: 26 September 2016 10:02 PM
             Updated: 26 September 2016 10:02 PM
                 Due: 28 September 2016 10:02 PM (2d 0h 0m)
      Resolution Due: 04 October 2016 10:02 PM (8d 0h 0m)



CONVL convolves images, usu. with Gaussians.    Although most images with units of Jy/beam are in fact also convolved with such a beam, its perfectly possible to have an image with units of Jy/beam which has *not* been convolved with any convolving beam - essentially the output of IMMOD is usually in this form.  This causes problems for CONVL

CONVL tries to convolve not *with* the specified BMAJ,BMIN,BPA, but rather it takes the convolving beam of the input image and tries to convolve *to* that specified size.  This doesn't work in the above case, where the
input image is not in fact so convolved, even though it *is* in Jy/beam.

CONVL's messages are rather misleading here: it says, e.g. 
"CONVL1: Convolve with: MAJ:   0.0017 MIN:   0.0008 PA:  164.6"
Even in cases where it not convolving with anything like the that (in the limit of the desired output size
being very near the input, it basically doesn't convolve at all, but still *claims* to be).

I was tripped up because I wanted to convolve an IMMOD output image to my target convoling beam size, and
since my target size was very close to that in the CG table (but in my case marginally above), I got an output image that was essentially not convolved with anything! (I was further confused because I'd forgotten the CG 
table and had used PUTH to make BMAJ,BMIN very small for this purpose, but CONVL of course ignored that and went for the CG table).

I think it would help if CONVL told you the actual size it *is* convolving with, so you can see whether its in fact convolving anything - probably in addition to telling you the output convolving beam size. It should not say "CONVL1: Convolve with: MAJ:   ..." when it actually means "Convolve *to*: Maj..."

However, the thing I was trying to do is probably not uncommon - could CONVL get an option, say 'CMOD' to *ignore* the input convolving beam, and just convovle the input image with the specified Gaussian for the purpose of convoling model images?  The convenient thing in the 'CMOD' case would be for BMAJ,BMIN,BPA to default to the convolving beam values already in the header/CG table.


 

------------------------------------------------------
Staff CP:  https://help.nrao.edu/staff