[daip] cross-correlation of theoretical images

[Eric Greisen]

Juan-Carlos Algaba-Marcos wrote:
> To whom it may concern,
> 
> I am lately working on task CONVL with theoretical and observed images. Currently I want
> to obtain the cross-correlation of a theoretical model and the corresponding synthesized
> 'observed' one. (OPCODE='IMAC') However, when I try to do so I obtain the following error:
> 
> localh> CONVL1: CONFIL: UNEQUAL SPACINGS IN TWO MAPS
> localh> CONVL1:        IMAGE=  0.0000  0.0000 CONV. FN=  0.0000  0.0000
> localh> CONVL1: Destroyed MA image file: catno=     39 disk= 2
> localh> CONVL1: Purports to die of UNNATURAL causes
> localh> CONVL1: localhost    31DEC11 TST: Cpu=       0.0  Real=       0
> 
> As the first image is a model, I have removed the blanks with task REMAG and adjusted
> all values to match the observed image with PUTHEAD. Here are the headers of both input
> images:
> 
>
IMHEADER does not display the axis increments with enough accuracy to 
see what CONVL is complaining about.  It checks that the increments are 
the same to 5% and they are not.  Try keyword='cdelt1'; gethead on one 
image and then puthead on the other.  Repeat with keyword='cdelt2'.
That should get us past the above error.

It is true that when you convolve with a Gaussian is scales correctly to 
produce Jy/beam from Jy/pixel and it should therefore change the header 
units.  keyword='bunit'; keystr='jy/beam'; puth will fix that.  I will 
look at correcting the code (in 31DEC12 version).

OHGEO is intended to interpolate an image to the coordinates of the 2nd 
image - your two are half-way across the sky from each other and so that 
is a meaningless operation (all blanks out results).

Eric Greisen