Convol and masked images

[and460]

Hi,

I've been using convol to try to convolve mosaics that I've been making with CABB data at 8MHz intervals to a uniform resolution throughout the 16cm band. I have noticed that on the convol documentation though, it says "WARNING: Convol does not do a very good job with masked images."

What is meant here by masked images? My mosaics have been created by feeding invert all the pointing data for the mosaic (i.e. the so-called joint approach), then running mossdi and restor on them. Around the edges of the mosaic at the limit of the primary beam, the field is blanked out to zero (masked?). Does this mean that convol will not produce good results for this application? What form does the 'not doing a good job on masked images' take i.e. incorrect flux normalisation etc?

Cheers,

Craig.

[Mark.Wieringa]

Hi Craig,

I think this warning is meant to remind you of the fact that convol uses FFTs to do the convolution. This means that any sharp edges (emission or noise cut off by the mask) will result in artefacts (ringing etc) in the image. It should be pretty obvious from the resulting image if this is happening in your case.

If there is little emission near the mask edge all should be well.

Cheers,

Mark

[and460]

OK - thanks for that Mark.

Cheers,

Craig.