[ Basic Info | References | User Guide ]
Basic Information on conterr
Task: conterr
Purpose: Continuum subtraction error estimate, when using UVLIN.
Categories: uv analysis
Given an image of the continuum in a spectral line experiment,
CONTERR estimates the residual continuum artifacts that will
be present after continuum subtraction using UVLIN. The results
are intended to be indicative, rather than rigorously accurate.
The peak value and location of the errors are approximately
correct, but the actual error pattern will differ significantly
from that presented here.
This task essentially just evaluates the expressions given
in Table 1 of Bob Sault's paper ``An analysis of visibility-based
continuum subtraction''.
CONTSEN makes plots of noise amplification factor of tasks such
as UVLIN, AVMATHS and CONTSUB.
Key: in
An extimate of the continuum. It should be either a dirty image
or a deconvolved/restored image. No default.
Key: order
The order of the polynomial fit in UVLIN. Either 1 or 2 numbers
can be given. If two numbers are given, then CONTSEN determines
the error for order(1) to order(2) inclusive. The
default is to determine the noise sensitivity for first order
only.
Key: bw
Total spectral bandwidth, in GHz. Default is 0.008 (i.e.
8 MHz).
Key: fwhm
The image resolution, given as the FWHM of the corresponding
gaussian, in arcsec. This is the same as the FWHM that
RESTOR would determine. The default is to get the information from
the input image. Two values can be given, being the FWHM in
x and y respectively (i.e. the position angle is assumed to
be along one of these axes). If only one value is given, the
gaussian is assumed to be circularly symmetric.
Note that high order fits are very sensitive to this parmeter.
Key: out
An output image containing the error images. The default is
no output image.
Key: options
Extra processing options. Several can be given, separated by
commas. Minimum match is used. Possible values are:
shift Determine an optimum shift of the phase centre to
be applied before the error is evaluated. This
shift will minimise the error in some sense.