Restoring an Image

or mfclean , you will probably want to use restor to convolve this with a Gaussian CLEAN beam (restoring beam) and add the remaining residuals. For models produced by mfclean , restor convolves only the intensity ( I ) plane with the CLEAN beam, although it does use the plane when determining the residuals.

Having the restoration as a separate step does have some flexibility advantage, but the user then has to run another task to produce the final image. Task restor

optionally fits the dirty beam with a Gaussian CLEAN beam, and then either produces a restored image or, alternatively, a residual image. There are a few other possibilities, but these are less common. The inputs to restor are:

• model: This is normally the output CLEAN component image, but can be any image that is a good representation of the source. It should have units of Jy/pixel.

• map, beam: These are the input dirty image and beam.

• mode: This parameter determines the operation to be performed by restor . The most common operations are to produce a restored image (mode=clean) or a residual image (mode=residual).

• out: The output restored or residual image.

• fwhm, pa: Parameters of the CLEAN beam; fwhm gives the size, in arcseconds, of the Gaussian beam to be used in the restoration. This will normally be two numbers, giving the full-width at half maximum of the major and minor axes of the Gaussian. If only one value is given, then the Gaussian is assumed to be circularly symmetric. Parameter pa gives the position angle, in degrees, of the restoring beam, measured east from north. If no values are given for fwhm, restor first checks the beam dataset to see if it contains Gaussian fit parameters. If not, restor performs a fit to the beam data. Note that the fitting procedure will probably give different values to the AIPS MX and APCLN tasks. Generally the value computed by restor is to be preferred to the APCLN and MX values.

Typical inputs are given below: